In this week’s AI x Gaming newsletter, we look into China’s pragmatic approach to applying AI in game development, pushing beyond efficiency to explore AI as a major opportunity to reshape its competitive position.

Also, a final reminder that the Naavik team will attend Gamescom on August 26 and 27. Reach out if you’d like to meet, discuss the AI ecosystem, or explore working with Naavik.

Sponsor: Powering the Future of Gaming with AI

Only 6.5% of playtime goes to new titles, even as the gaming industry grows 7.5% a year. Retention decides who wins.

The next generation of gaming will be powered by intelligence, not just content.

The problem: player data is scattered across telemetry, social, transactions, storefronts, support, and platform identities. Most studios struggle to answer the basics — who is this player, what do they care about, and how do we grow their value? Fragmented systems make it difficult to create a complete view of each player.

AI-driven Player 360 transforms this data into real-time intelligence, helping studios understand player behavior, predict churn, personalize experiences and optimize monetization. By integrating unified data, cross-platform identity resolution, scalable AI models and real-time analytics into the core infrastructure, studios can move from simply understanding what happened to anticipating what happens next. It can even trigger in-game decisions, turning prediction into a personalized, in-session offer.

The result? More adaptive gameplay, smarter engagement, stronger retention, higher lifetime value, and greater operational efficiency.

Discover how AI-driven Player 360 is powering the next generation of gaming!

Learn More

China's Big Bet on AI for Games

Written by Mark Ma, Naavik Contributor and Principal Strategist at JM APAC Strategies

I recently tried an AI-driven detective puzzle game that an independent Chinese developer released on itch.io. The concept is simple enough: playing the role of a detective, you interrogate suspects powered by a large language model that allows them to improvise in character with spoken audio to keep you off balance, requiring you to sift through their answers to solve the crime. The visual presentation still feels a bit rough, and the core mechanic comes across as incremental rather than groundbreaking. Yet, it signals a broader shift: in China, integrating AI directly into core gameplay has evolved beyond isolated experiments; it is now a systemic trend.

Ever since generative AI emerged, Western game developers have primarily focused on copyright concerns, ethical questions, and fears over job displacement. In contrast, studios in China pivoted to a more pragmatic question: how can AI drive practical results? Today, it’s clear that AI is already helping Chinese developers reduce operational costs and streamline production. Yet rather than stopping at efficiency, they continue to push boundaries, investing heavily in unproven, forward-looking applications.

A Tale of Two Speeds

Whenever AI is brought up, a sense of lingering anxiety seems to shadow the Western gaming scene (more so than other industries). The 2026 GDC State of the Game Industry survey found 52% of developers think generative AI is hurting the industry — up from 30% a year earlier and 18% the year before that. As that sentiment has taken hold, AI’s rollout in the West has faced noticeably higher resistance. Most notably, Steam makes developers disclose AI use, and more than 7,300 games had a label as of March. Larian promised to drop generative AI from the concept-art phase of its next game after fan backlash. Call of Duty: Black Ops 7 caught complaints over its AI-made calling cards too. AI still slips into backend work — EA’s Battlefield team runs LLMs to catch duplicate Jira tickets, for instance — but the moment the output lands in front of players, the backlash gets noticeably louder. Many companies are bending the knee as a result.

In China – not unlike mature Western markets – rising development and user acquisition expenses are squeezing margins, and hit rates are falling as companies increasingly compete for existing audiences. However, in response to these pressures, China’s game industry is moving at a completely different pace.

Looking at distribution, Tanwan Games reports an AI adoption rate surpassing 80% across art creation and ad targeting. Similarly, 37 Interactive’s 2024 annual report indicates that AI produces over 80% of its 2D art and assists in more than 70% of its ad video creation.

In core gameplay, AI integration is already directly embedded into major titles. NetEase’s Justice Online markets its dynamic, AI-driven NPCs as creating “a Jianghu that breathes,” while Giant Network’s Supernatural Squad recorded over 25M matches against AI opponents during its launch week.

When it comes to art pipelines, Tencent’s VISVISE platform stands out. Character rigging traditionally took one to three and a half days, while generating a ten-second skeletal animation required three to seven days. VISVISE accelerates character rigging by more than 8x and has been deployed across nearly 100 projects, including flagship titles like Honor of Kings, League of Legends: Wild Rift, PUBG Mobile, and Teamfight Tactics. Beyond a mere feature set, its true strength lies in execution: developed hand-in-hand with the PUBG Mobile production team, the core GoSkinning module (v4.2) was tailored to overcome practical hurdles such as skirt physics and four-legged rigging. Such close integration with an active pipeline provides a competitive advantage that off-the-shelf software would struggle to match.

Is Moving Faster with AI Paying Off?

Gamma Data estimates AI could add ¥53.3–84.6B (roughly $7.9–12.6B) in new market value for China’s game industry. Net profit across the industry is projected to grow from roughly ¥111.37B ($16.5B) in 2026 to ¥162.66B ($24.1B) by 2030, with AI as one of the major drivers.

Among China’s top 50 game companies by revenue, 80% have disclosed AI plans publicly, and 56% have built dedicated labs or teams. A 2025 Niko Partners survey puts AI usage at around 60% of Chinese studios. In the first half of 2026, AI-driven titles were just 15% of China’s top 100-grossing mobile games, but they generated 42.6% of the revenue. This also reflects strong products and mature teams that are better positioned to leverage AI, rather than just newer entrants.

However, we’re also starting to see examples of new games that benefit from more efficient AI workflows. For example, 37 Interactive’s AI-native mini-game, Heart of the Moon Palace, would normally have required a development cycle of six months or more. But with AI helping drive the story and art, a team of just two or three people shipped it in about 10 days instead.

Clearly, AI adoption is not just a future promise. It is actively driving revenue, reducing costs, and expanding user engagement. In a landscape where rapid execution is critical, China’s gaming sector has responded by swiftly scaling and industrializing AI workflows.

AI Industrialization — Then What?

When AI adoption becomes universal, and its capabilities are widely understood, how do teams gain a true competitive edge? miHoYo’s strategy illustrates one extraordinarily bold approach to this dilemma.

miHoYo co-founder Liu Wei said in May that the company plans to invest up to ¥100B (roughly $14.8B) over the next three years, building its own training systems, GPU clusters, and models from scratch, with NPC intelligence and content generation as the main targets. An early test case is Petit Planet, a life-sim title currently in beta testing designed with persistent-memory NPCs that adapt their behavior based on past interactions. For instance, helping a villager find a misplaced toolbox earns the player pest-resistant seeds the following day, whereas declining to help results in higher tool prices down the line. He didn’t dress it up: “Even if it doesn’t work out in the end, we accept that. Think of it as fireworks.”

That’s a lot of money! For a pure-play gaming company, it’s a huge bet worth at least two years of company profits. As such, it also represents genuine risk. A 2025 survey from the China Audio-Video and Digital Publishing Association’s game committee found 22.7% of companies think AI investment isn’t worth the cost; MIT research from the same year found 95% of enterprise generative-AI projects showed no measurable financial return within six months. Liu Wei’s “fireworks” line suggests that the company understands the risk, but that doesn’t make the risk any less material.

Leading Western publishers tend to treat uncertainty as risk to be minimized. Chinese companies, in contrast, are more likely to view missing a major opportunity as the larger danger. This gambling instinct — moving before there’s any certainty — isn’t unique to gaming either. In February and March, the open-source AI agent framework OpenClaw, nicknamed “the lobster,” set off a nationwide craze; by May it had clearly cooled off. The hype outran its own commercial and safety capabilities.

We can spot this urgency in the hiring data. 51job’s 2026 Online Game Talent Insights ranks “AI algorithm/machine learning engineer” and “AI-generated content (AIGC) application engineer” fourth and fifth among the most in-demand roles, both ahead of the traditional “game development engineer” at eighth. The report also found that companies outside the state-owned sector now account for nearly 80% of all hiring across the industry and notes that these firms tend to respond faster to shifts like AIGC. None of this is unique to China, but while much of the world still treats AI in gaming as optional, China’s games industry is already well past the point of no return.

While there are benefits to larger companies like miHoYo (such as redefining how AI-native gameplay and production work at scale), leveraging AI unlocks material advantages for smaller studios and indie developers, who can do more than ever before. Consider a developer who shared a breakdown this May: in just two months of off-hours work and a budget of ¥1,200, they built a complete WeChat mini-game. By dividing the workload among AI systems — using Claude Code for programming, Claude for text, ElevenLabs and Pixabay for audio, and GPT, Gemini, and Doubao for art — a single person handled roles that previously required an entire multidisciplinary team. The end result featured endless levels, friend leaderboards, and daily quests, even if its commercial viability remains unproven. While this phenomenon extends beyond China, the country stands apart due to its sheer volume of developers and their willingness to jump in despite the uncertainty.

Conclusion

Is this momentum likely to continue? Almost certainly. Facing a deceleration, China’s gaming sector has found in AI a major opportunity to reshape its competitive position.

Whether this translates into accelerating market share gains globally remains undetermined, but it would be a continuation of what’s already been happening for years. Currently, AI provides the greatest leverage in mobile development and indie projects (areas where Chinese studios are already well established), but in the long run, expensive AAA productions face the largest competitive threat from business model disruption. I expect a surge in Chinese game production: faster output and broader participation, though not necessarily higher success rates.

Western observers must realize that there is no turning back. Reluctance toward AI in one region will not stall advancement elsewhere. Disparities in adoption speed and openness to innovation will ultimately carry competitive consequences, even if the transition involves missteps or excessive investment. Because the Chinese gaming market regards AI adoption as inevitable, that expectation becomes self-fulfilling, leaving open the question of when Western counterparts will come to the same realization. Of course, not all of the West is entirely lagging (mobile is seeing faster innovation, not to mention the U.S. leads the world in AI research), but console and PC gaming markets in particular are moving at much slower speeds.

As China’s game industry moves forward aggressively, an influx of developers will continue experimenting with these technologies. It remains to be seen to what extent Chinese studios can leverage AI to create novel gameplay experiences — an area where progress currently lags — but their active AI-enabled development at scale will gradually pull the global industry in the same direction.

Top News Items

Written by Devin Becker, Consultant at Naavik

Unity Records "Best Quarter in History," Powered by AI Advertising

• Unity posted $546M in Q2 2026 revenue, up 24% YoY, and CEO Matthew Bromberg called it the company’s best quarter ever — but not due to its game engine. The Grow division, powered by Vector AI, its AI-driven advertising platform, surged 69%. Net losses narrowed from $108M a year ago to just $23M, and the stock jumped roughly 20%. Bromberg framed the business as a flywheel: as more developers build on Unity and more players engage with those games, more data trains the ad platform, and the ad platform gets better at serving games to players. It’s an ecosystem pitch, and investors seem to be buying it. Unity continues making improvements to Vector AI that helped drive growth, including rapid adoption of its Day 28 ROAS campaign tracking, increasing agentic capabilities, ingestion of legacy ironSource data, and hybrid monetization optimizations.

• As Unity pushes deeper into the ad platform business, there’s a question of how big a competitor it can become compared with others like AppLovin. There’s currently a fairly big gap, but Vector is on a strong trajectory with four straight quarters of accelerating sequential growth, and the runtime data integration gives Unity a structural advantage that other big players can’t easily replicate. Unity’s unique position as the engine powering the majority of top mobile games means it can see behavioral data from inside the game session itself, not just around the ad impression. If that runtime signal meaningfully improves prediction accuracy in combination with increasing automation, it could significantly increase Vector’s competitive position. Catching AppLovin is unlikely any time soon, but becoming a strong second option in a market where developers are actively worried about AppLovin’s dominance is a possibility if improvements continue at this rate.

Gamescom Dev Survey: Developers Expect AI to Reshape Teams Within Three Years

• A survey of 100 speakers at this year’s Gamescom Dev Conference found that 83% expect generative AI to affect team structure or productivity in the near term, but the consensus fractures from there. 36% believe AI will change roles without reducing headcount, 33% expect smaller teams, 17% don’t expect team reductions at all, and only 14% anticipate higher individual output. In terms of current use, code and production led at 34%, while 30% of respondents said they want as little AI in game development as possible. Art, animation, narrative, and localization collectively barely registered. These results are broadly consistent with the same survey from a year earlier, suggesting the industry’s ambivalence has stabilized rather than resolved.

• Developers are converging on the idea that AI will matter but cannot agree on how it will matter, what it will change, or whether the change is desirable. Meanwhile, as discussed above, Chinese studios and publishers are making aggressive, top-down bets on AI integration across production pipelines. The gap between Western hesitancy and Eastern acceleration is widening. Studios that spend the next few years arguing about acceptable use policies while competitors ship AI-assisted games faster and cheaper are ceding ground. The backlash cycle of “use, get caught, apologize” will fade as AI-assisted production simply becomes production.

AI Cloning Is Becoming an Industrial-Scale Threat to Indie Developers

• Indie developer Freya Holmér posted a 50-second prototype clip of a rotating Tetris concept, and within days multiple vibe-coded clones appeared on app stores. One cloner admitted that the process took about a day using AI tools, but the problem goes beyond solo copycats. A former employee of Midnight Works admitted that the studio’s established playbook of cloning popular games and using confusingly similar titles has now been accelerated by AI at every step, from banner art to 3D models. Game cloning is as old as the industry itself, but AI collapses the time and cost of producing them. Clones now proliferate not just on mobile app stores but across UGC platforms like Roblox, where the barrier to publishing is even lower.

• For indie developers, the instinct is to share less publicly or to treat this as an existential crisis, but ultimately both are counterproductive. The harder truth is that a novel game concept was never a durable competitive advantage on its own. Ideas getting copied and built upon is part of the way games evolve. What separates successful indie games from their clones is everything that surrounds the idea, including superior execution, community building, smarter marketing and distribution, with the kind of polish and personality that a vibe-coded knockoff can’t easily replicate. Developers investing in those layers are building moats that resist cloning regardless of how fast AI tools get. The teams that treat a cool prototype as the product rather than the starting point are the ones most exposed here.

100% of Japanese Online Game Developers Are Using AI, Mostly for Player Analytics

• A report from Japan’s Online Game Association (JOGA) and Kadokawa ASCII Laboratories found that 100% of surveyed Japanese online game developers are using generative AI in some capacity, with some important caveats. The dominant use cases are user preference analysis and user behavior prediction, not content creation. Google Gemini led with a 94% usage rate, followed by Claude at 84% and GitHub Copilot at 76%. Japanese players, meanwhile, expressed concerns about AI making games look and feel too similar to each other.

• The Japan data reveals a split that the Western discourse often misses. Much of the actual AI adoption in game studios, particularly those running live-service titles, is happening in analytics, player modeling, and operations rather than in visible creative work. This makes sense as online games live and die on retention, and AI-powered behavior prediction is a direct lever on that metric. The players’ concern about homogeneity is probably overstated as developers and the market will still highly value creativity. This survey shows that Japan, unlike the West, isn’t hesitant to adopt AI but is doing so where it has the best ROI and the least creative risk.

Fenris Creations Hires Former Icelandic Minister to Lead AI Research Initiative

• Fenris Creations (formerly CCP Games) hired Áslaug Arna Sigurbjörnsdóttir, Iceland’s former minister of higher education, science, and innovation, as senior director of AI partnerships. The role is purpose-built: Sigurbjörnsdóttir will lead partnerships with AI labs, universities, and policymakers around a research initiative that positions EVE Frontier, the survival spinoff of EVE Online, as a live proving ground for autonomous AI agents (as much as it was for blockchain tech). The premise is that EVE’s persistent, player-driven economy and its unpredictable social dynamics offer something short-lived simulations cannot, with its realistic, ongoing environment in which AI agents can be studied alongside real people making real decisions. The initiative builds on Google DeepMind’s earlier minority investment in the studio.

• Sigurbjörnsdóttir’s specific job is to expand the research partnership model beyond DeepMind, bringing in additional academics, researchers, and policymakers. The DeepMind deal already helped finance Fenris’s return to independence, and more partnerships could bring additional funding and strategic relevance to a studio that recently returned to profitability in 2025. No other game studio is seriously attempting to position a live MMO as formal AI research infrastructure at this scale. If it proves useful for studying long-horizon planning, adversarial memory, and multi-agent decision-making, other developers with complex persistent worlds may try to make deals of their own.

Other News Items

• Saber Interactive to add Rideshare ‘Stimulator’ AI disclosure after public controversy

• AI startup Reforged Labs shuts down as the “gap we were selling into is closing”

• How AI, gaming, and college workloads are changing student laptops

• Amazon draws on Assassin’s Creed lore for its first fully AI-generated television series

• Twitch will sacrifice you to its AI overlord whether you like it or not

• Redditors’ Fake Duskbloods Boss Guides Fool Google’s Dumbass Delusion Machine

• Europe’s AI labeling and transparency rules are now in effect

• Autodesk launches 3D Editor + Canvas to bring generative AI to filmmaking

• The major labels propose rules to keep AI slop off the charts

Content Worth Consuming

• From TikTok to ‘mini apps’: How Sekai wants to transform vibe coding into a creator medium (Gamesbeat): “For now, Sekai does not have a full-fledged creator economy. Creators can build and share apps on the platform, but can’t directly monetize their creations on the platform. Zhang said his team is not currently focused on monetization, viewing it as something that will naturally follow once creators build momentum on the platform.”

• Four takeaways from Mark Zuckerberg’s massive AI manifesto (The Verge): “Doing so ‘will bring an abundance of personal agency, expression of ideas, deepening of relationships, economic and financial prosperity, improved health, and inventions we cannot imagine today,’ according to Zuckerberg. He also says this will have an impact on ‘some aspects of the way we work,’ but waves off concerns that AI will result in fewer jobs overall.”

• Creative as the Only Lever You Control, Attribution as a Hunt for the Truth, and the Death of the UA Operator (Sett): “In this episode, Amit sits down with Amos Adler, VP of Growth at SciPlay and a graduate of the Dawap Mafia, the Israeli mobile marketing school that produced a large chunk of the people running UA today… Amos talks about growth the way an analyst talks about a system: creative as the only lever fully in his hands, attribution as a permanent hunt for the truth rather than a one-time decision, and a near future where the UA operator role disappears and commerce, CPA and IAA run 90% autonomous.”

• The battle over Tung Tung Tung Sahur is testing the limits of copyright and trademark law (Gamesbeat): “As the legal dispute over AI-generated meme ownership plays out, both sides of the case are concerned over the optics of the case, with both the plaintiff and defendant positioning themselves as warriors in a broader moral crusade. Moss framed the case as a battle to prevent Mementum and other entities from stifling creators’ use of AI-generated memes, which he described as developed and owned by the public; Stein described Mementum’s position as fighting for all creators’ ability to assert and monetize the rights to their content.”

• Steam Is Crushing AI Slop (YouTube): “Devs told Game Discover Co. Steam is checking, Steam is catching it. And in one case, Steam was actually able to pick up that a background artist had used some GenAI stuff for their art when the actual dev didn’t even notice themselves. And that means that these Steam disclosure numbers actually do mean something. And the scenario they point to is that by 2028, 50% of all games on Steam could have an AI disclosure. That’s a big number.” Source analysis here.

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the game industry. We’d love to work with you too! Here’s how we can help, spanning all platforms, genres, and regions:

• Strategy Consulting: Market research, corporate strategy, game and economy design, gamification, live-ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above to learn more. If you'd like to discuss how Naavik can support your team, click the box below or send us a note at consulting@naavik.co.

First, a quick reminder: Naavik will be at Gamescom this year! If you’d like to connect and learn more about how Naavik can support your team, please reach out here.

Moving onto our newsletter — in our inaugural issue, we deeply analyzed mobile F2P’s H1 2026 performance and what it means for the future of the industry. In today’s issue, we do something similar, but instead take a closer look at one of mobile gaming’s most enduring subgenres: 4X Strategy.

Sponsor: How Highway Racer Pro Blew Up Organically

Highway Racer Pro hit #1 in the U.S. with zero paid marketing. When launch day traffic surged past forecast, Heroic Labs’ engineers handled troubleshooting, allowing Spektra’s small team to focus on the racing their game deserves, not worrying about whether the backend can hold.

Learn More

H1 2026: 4X Strategy’s UA Reckoning

Written by Max Abrahamsen, Consultant at Naavik, and Abhimanyu Kumar, Naavik’s Co-founder

Compared to mobile F2P and its parent genre, Strategy, 4X Strategy’s H1 2026 performance looks impressive. 4X Strategy downloads rose by +5% YoY (+18M), while total mobile F2P fell by -12% YoY (-3B) and Strategy by -3% YoY (-29M). IAP revenue was also up +5% YoY (+$193M), compared with declines of -2% YoY (-$0.6B) for mobile F2P and -5% YoY (-$319M) for Strategy.

However, these six-month totals conceal a sharp quarter-over-quarter decline within H1 2026. The subgenre peaked during Q4 2025 and has since declined, with H1 2026’s YoY growth being concentrated in Q1. During Q2 2026, 4X Strategy’s downloads fell -27M QoQ (-12%) versus -5% across mobile F2P and -9% for the Strategy genre. Revenue declined similarly, dropping -$195M QoQ (-10%) versus -6% across mobile F2P and -11% for the Strategy genre. Notably, 4X Strategy accounted for approximately half of the Strategy genre’s revenue decline.

Examining the potential reasons for the deterioration of mobile F2P’s largest revenue-generating subgenre over H1 2026 is where 4X Strategy’s analysis gets interesting. Let’s jump in.

Data caveat: Throughout this article, data excludes DTC, China Android, and ad revenues. Also, the market size figures come from Sensor Tower’s “Market Size” feature, which covers a subset of the entire mobile gaming market. While the aggregate numbers are smaller than the broader mobile gaming market, we’ve confirmed that this subset is representative of broader market trends. We’ve also flagged where to keep the missing DTC revenue in mind when studying broader subgenre trends.

4X Strategy’s Slowdown over H1 2026

4X Strategy’s +18M YoY downloads increase during H1 2026 was not really concentrated on a particular platform or region. The relatively minor +5% YoY growth hints towards substantial title-level movement beneath the surface. Roughly 40% of all tracked 4X Strategy titles grew at various intensities, adding +150M YoY downloads. Around 50% saw small to large download drops, collectively driving a -131M YoY decline. The remaining 10% were flat. However, what’s interesting to note is what kinds of titles drove growth and which declined.

When slicing the data by title launch year, it was primarily new 4X Strategy launches over the last two years that added +112M downloads and drove the subgenre’s H1 2026 downloads growth. This included titles like X-Clash (+26M), Kingshot (+18M), Tiles Survive! (+17M), Last Light (+15M), Last Asylum (+14M), and more. Some of these games launched after H1 2025 (adding downloads to H1 2026 versus zero contribution to H1 2025), while others continued to scale UA over H1 2026.

On the other hand, 4X Strategy’s older portfolio (pre-H2 2024) offset -94M downloads of that growth. This included games like Lords Mobile (-13M), Total Battle (-11M), Dark War: Survival (-10M), Last War: Survival (-8M), Evony (-8M), among others. Some of these games are long-standing incumbents that are naturally unable to scale UA to the same degree as newcomers, while others scaled rapidly during 2025 but could not sustain that pace into 2026.

On the revenue side, ~30% of all tracked 4X Strategy titles grew at various intensities over H1 2026, adding +$962M in YoY IAP revenue. Around 45% declined, collectively removing -$769M, while the remaining ~25% were flat. The revenue story was similar to downloads; the subgenre’s older portfolio absorbed ~70% of the newer games’ revenue growth, leaving 4X Strategy up +$193M YoY over H1 2026.

More specifically, newer titles released over the last two years added +$620M YoY in H1 2026, while older titles took away much of that growth (-$425M YoY). Kingshot was the major growth contributor (+$403M), given its early 2025 launch and rapid UA scaling since. Last Z was another bright spot (+$193M YoY) — it technically belongs to our definition of the older portfolio due to its mid-2024 release, although it truly began scaling over 2025, and the game’s H1 2026 results inherited that marketing momentum. However, established incumbents declined significantly. Last War: Survival saw steep declines in downloads, MAU, and ARPDAU over H1 2026, resulting in a significant -$135M YoY (-17%) decline. Whiteout Survival was also down -$126M YoY (-16%) due to declining MAUs, primarily driven by falling downloads.

Based on the graphs and notes above, two observations emerge:

1. 4X Strategy’s performance would’ve been flat to declining if not for newly released titles that successfully scaled UA during their launch years, thereby driving up overall subgenre revenue.

2. For older 4X Strategy games, sustaining revenue seems to require regular top-of-the-funnel replenishment through stable new- and returning-player-focused UA strategies — especially when new competitors appear. Keeping long-term retention healthy and improving ARPDAU could even result in growth.

The above represents why we found 4X Strategy’s downloads and revenue deterioration over H1 2026 surprising. As a reminder, between Q1 and Q2 2026, 4X Strategy’s downloads fell by -27M (-12% QoQ) and revenue by -$195M (-10% QoQ).

On performing a revenue-weighted analysis of which metrics drove 4X Strategy’s revenue decline over H1 2026, a broader DAU deterioration drove ~73% of it, ~92% of which was attributable to new user acquisition and ~8% came from natural erosion of existing users. The remaining ~27% was attributable to lower ARPDAU over the period.

This demonstrates the critical role of UA in sustaining overall 4X Strategy health — and over H1 2026, UA didn’t play that role very well. Among the subgenre’s top 100 U.S.-grossing games over H1 2026 (representing ~98% of U.S. subgenre revenue), 81 games were running UA in 2025, and 68% of them recorded YoY declines of at least -10% in their U.S. UA channel impression share. Quarter over quarter, 8 of the top 10 grossing saw a similar intensity of impression share declines in the US. Further, 54% of all new releases, and 60% of all old games also recorded impression share declines of at least -10% in the US.

While this does not mean that the top 100 grossing 4X Strategy games stopped acquiring users altogether, it indicates that many of them were no longer investing in UA with the same intensity as before. This could be an early and critical indicator that 4X Strategy’s ability to acquire users with its current UA tactics is beginning to weaken in a new UA environment that’s rearing its head more prominently over 2026.

Why 4X Strategy Might Be Pulling Back on UA in H1 2026

There is likely no single explanation for this UA pullback over H1 2026, but here are a few factors we believe could be contributing.

#1 – A Contracting Mobile Downloads Market: Mobile downloads fell by -12% YoY during H1 2026, despite gaming ad spend increasing by +8% and impressions by +14%. Publishers likely spent more to compete for a shrinking pool of installs, putting pressure on acquisition costs and weakened returns. For top-grossing 4X Strategy titles, such UA conditions could’ve made previous levels of marketing investment increasingly difficult to justify.

#2 – Midcore Impression Share is Increasingly Consolidating: As mentioned in our H1 2026 mobile F2P analysis, AppLovin and AdMob now control roughly two-thirds of mobile gaming’s ad revenue. This percentage moves up to ~75% when looking only at midcore genres, and consolidation accelerated over H1 2026. 4X Strategy lives within and also acquires new users more broadly from midcore subgenres. Therefore, as midcore’s ad impression share increasingly consolidates around fewer networks, 4X Strategy publishers increasingly compete within a more concentrated auction pool — not only against one another, but also against advertisers across other midcore subgenres. That, of course, could put more pressure on 4X Strategy’s H1 2026 UA economics.

#3 ‒ Non-midcore Subgenres are Competing with 4X Strategy for Impression Share Too: The point above also extends to non-midcore subgenres, whose impression shares are increasingly consolidating towards AppLovin and AdMob too. This can be seen in the graph below, where non-midcore subgenres like Match Merge 2 and Sort have contributed to 4X Strategy’s declining impression share over 2026, likely driven by rapidly scaling games like Gossip Harbor (Match Merge 2) and Pixel Flow (Sort).

Taken together, these market factors could’ve created tougher UA conditions for 4X Strategy — a subgenre that has historically relied on UA for stability and growth. While both new and old games seem to be impacted, new titles at least have a novelty factor to combat the evolving UA environment. On the other hand, older games have already spent years running UA profitably at scale. For them, as the most receptive audiences are gradually acquired and re-targeted, reaching additional high-value players can become increasingly difficult. Therefore, older titles might generate better returns by prioritizing retention and monetization improvements.

That said, acquisition is only the beginning of 4X Strategy’s modern commercial funnel, which we covered here. Broadly appealing advertising and highly accessible onboarding must ultimately convert players into the deeper 4X experience. Pressure at the acquisition stage of the funnel will therefore affect each game differently depending on how well the game has executed on the stages that follow.

How 4X Games are Reacting to Harder UA in H1 2026

This part of the analysis will focus on the four 4X Strategy titles that clearly dominated the subgenre’s revenue ranks over H1 2026 — Last War: Survival, Whiteout Survival, Kingshot, and Last Z. Interestingly enough, each of these titles seems to be tackling 2026’s UA environment slightly differently.

#1 ‒ Last War Might be Retrenching: No title contributed more to 4X Strategy’s Q2 2026 contraction than Last War: Survival. IAP revenue fell from ~$406M in Q1 to ~$278M in Q2 (-32% QoQ), accounting for approximately two-thirds of the subgenre’s quarterly revenue decline. This has primarily been a function of the game taking two very questionable steps in 2026:

1. UA seems to have been effectively turned off since May 2026. The game reported near-zero impression share on Sensor Tower over the last three months, with slow UA reductions since January 2026. This, combined with natural existing audience erosion, impacted the game’s MAU curve, collectively driving ~72% of the game’s quarterly revenue decline over 2026. MAUs might settle at a core audience baseline over time.

2. The developer also reduced the game’s monetization pressure in April 2026 by removing a broad selection of purchase opportunities, which impacted the game’s ARPDAU baseline. This move coincided with a Discord message titled “Returning to Our Roots”, where the developer informed players that they agreed with player feedback around how “the constant pace of spending has started to feel overwhelming”, and took action by making “adjustments to some packs, events, and certain purchase entry points, as well as how they are presented in the game”. While this move might’ve been spurred by the game’s ARPDAU gradually declining over 2025 (as the active user base grew), the developer ended the letter in a very telling way: “We know this kind of decision might be confusing. It goes against the usual logic of this industry, and it means accepting some very real costs on our side. Some people may not understand it. Some may feel that these are only words. We understand that some people may see it that way. But if we don’t take even this step, then talking about returning to our roots means nothing at all.”

While we cannot be absolutely certain why Last War: Survival is going down this seemingly self-sabotaging route after dominating the subgenre since 2024, a possible (optimistic) hypothesis is that the game might simply be retrenching – scaling back UA to potentially wait out a harder UA environment, doubling down on improving retention of its most valuable players in the meantime, and evolving the monetization strategy in a way that fits the lifecycle stage of the game. After all, not many 4X Strategy games last forever, and 2026’s UA environment might be making Last War: Survival realize that truth even more. If the hypothesis is true, the key question for Last War: Survival will be whether this approach can minimize established audience churn without sacrificing too much revenue in the process.

#2 ‒ Last Z is Cautiously Returning to Acquisition: Last Z followed a similar UA playbook to Last War during its initial scaling phase. But starting from Q4 2025, ad impression share started to weaken. Interestingly enough, Q4 2025 was also the time Last War: Survival made a major (and possibly last) UA push, which could’ve impacted Last Z’s impression share figures. Fortunately, this Q4 2025 movement did not bubble up to the game’s downloads and MAU curves. However, during 2026, Last Z significantly scaled back UA efforts, which drove half of the game’s quarterly revenue decline. The other half was driven by lower ARPDAU in Q2 2026, which is also when Last Z cautiously started turning UA back on.

The game’s new UA strategy seems to revolve around reducing its focus on AppLovin and AdMob, and instead trying out new channels like TikTok and Moloco. However, the game’s creative strategy hasn’t changed much. Overall, it seems like Last Z has simply chosen to replenish its lost user base by trying a new channel mix. Although given the ARPDAU weakness coinciding with increased UA, it seems like the game might need to go back to the drawing board to find a new growth strategy.

#3 ‒ Century Games Focused on ARPDAU Improvements: Unlike Last War: Survival and Last Z, Century Games took a very different path with Whiteout Survival and Kingshot to combat weakening audience metrics over H1 2026.

Whiteout Survival’s quarterly revenue did not decline in H1 2026. It increased from ~$318M in Q1 to ~$325M in Q2 (+$7M, +2% QoQ). Under the hood, the game did see downloads deteriorate over the same period, and MAU followed. However, this new-user-driven decline was more than made up for with a very healthy ARPDAU over H1 2026. The game did not change much with its UA channel or creative mix either. In other words, if the game did not see healthy ARPDAU, its new user decline would’ve driven a roughly -$23M quarterly loss.

Kingshot showed a similar story over H1 2026. The game’s revenue increased from ~$236M in Q1 to ~$255M in Q2 (+$19M, +8% QoQ). This was primarily because the game’s downloads decline was more than made up for by healthy ARPDAU. If ARPDAU wasn’t healthy, the game would’ve seen a roughly -$20M quarterly revenue decline due to reduced new player supply. That said, the key difference in Kingshot’s story is that it is still in the early innings of its lifecycle, which means it’s likely getting significant backing from Century Games’ UA warchest. This is probably why it was also able to fly above H1 2026’s harder UA environment and maintain a relatively stable impression share throughout the period. It may also be why Whiteout Survival’s impression share dropped to a new baseline in 2025, when Kingshot launched and took the company’s spotlight.

Where Does 4X Strategy Go From Here

4X Strategy remains mobile gaming’s largest subgenre by IAP revenue (~15% share). But the subgenre’s deterioration over H1 2026 raises questions about the growth model behind it. While top 4X Strategy publishers seem to be reacting to the evolving UA environment in different ways, how long the current UA environment persists through 2026 and 2027 will determine how 4X Strategy will fuel its next stage of commercial growth.

That said, and on a more encouraging note, this is not the first time that changing market conditions have forced the subgenre to adapt. Since 2010, 4X Strategy has cycled through four generations. Each generation thrived under its market conditions while standing on the shoulders of the previous generation to strengthen the subgenre’s fundamentals. This can be seen in the image below, where the table showcases how Generation 4 utilizes all the goods from previous generations while adding something new on top.

While the subgenre is currently in Generation 4, the UA environment pressures visible in H1 2026 could be creating the conditions for an imminent Generation 5. If the consolidation of ad impression volume becomes more permanent, we believe 4X Strategy will need to:

• Focus its marketing efforts on either stronger differentiation or net-new user pool unlocks — both difficult endeavors, with the former likely being slightly easier than the latter.

• Focus its product efforts on finding new ways to either monetize highly valuable acquired users for longer periods of time or simply improve long-term retention — again, both tough nuts to crack; the latter is especially a long-standing untapped area of the subgenre, and we sometimes like to call it one of 4X’s last battlegrounds.

In other words, 4X Strategy’s era of acquisition at any cost may currently be facing some headwinds, but that may not necessarily be a long-term threat for a subgenre with a long history of successfully adapting to disruption. It may instead be the catalyst for its next evolution. What form this evolution takes remains uncertain, but innovation will likely be required across both marketing and product.

Top News Items

Written by Max Abrahamsen, Consultant at Naavik

Apple objects to UK regulator’s attempt to open up the App Store

• The UK Competition and Markets Authority has proposed allowing iOS developers to direct users outside their apps to complete digital purchases. Apple could still charge a fee on those transactions, but, according to regulators, the terms would need to be fair, reasonable, and make steering commercially viable for developers. Apple argues that these restrictions would amount to price regulation and give the CMA an intrusive role in determining how it monetizes the App Store. The consultation has closed, and the regulator is assessing responses before reaching a final decision.

• For mobile game publishers, the value of steering will be determined less by the right to link externally than by the fee Apple is allowed to charge. A sufficiently low fee could make webstore discounts worthwhile, improve margins, and give publishers more control over pricing and player relationships. If the fee remains close to Apple’s existing commission, those benefits could be absorbed by payment-processing costs and the added friction of moving players outside the app. The CMA is therefore tackling the most important weakness in similar reforms: payment choice means little when platform fees make the alternative commercially unattractive. Its final framework will determine whether external payments become a meaningful monetization route in the UK or mainly a formal compliance option. Similar disputes have played out internationally, most directly in the EU and U.S., while South Korea and Japan have adopted their own alternative payment regimes. Across these markets, the argument has increasingly shifted from whether alternatives must be permitted to whether Apple’s fees and conditions make them commercially viable.

AppLovin shares plunge nearly 20% as ad model delays raise growth concerns

• AppLovin reported Q2 revenue of $1.92B, up 53% YoY but slightly below expectations, while its Q3 outlook also disappointed. Management attributed the shortfall to a quarter with fewer meaningful improvements to its advertising models, saying the company “fell short of that standard.” However, a significant model upgrade launched shortly after the quarter ended, and management said Q3 started strongly. This suggests the weaker quarter may have reflected the timing of model improvements rather than softer advertiser demand.

• AppLovin remains one of the most important user acquisition and monetization platforms in mobile gaming, and 53% YoY revenue growth demonstrates the continued strength of its advertising business. The sharp share-price decline reflects the exceptionally high expectations embedded in AppLovin’s valuation more than a meaningful deterioration in its competitive position or product capabilities. Slower growth in gaming advertising and fewer model improvements are worth monitoring, but one weaker quarter provides little evidence that AppLovin’s competitive advantages have diminished. Such reactions are common when high-growth companies deliver strong results that nevertheless fall short of Wall Street’s expectations.

Krafton achieves record Q2 results with revenue up 95%

• Krafton delivered record second-quarter revenue and operating profit in Q2 2026, driven by the PUBG franchise and Subnautica 2. Revenue rose 95% YoY to ₩1.29T ($899.5M), while operating profit increased 67% to ₩410.9B ($286.5M). PC was the main growth engine, with revenue up 155% to ₩560.4B ($390.8M), helped by Subnautica 2 selling more than five million units within 22 days of its early access launch. Mobile remained Krafton’s second-largest platform, generating ₩451B ($314.5M), up 6% YoY, as PUBG Mobile and Battlegrounds Mobile India continued to grow.

• PUBG Mobile remains a durable asset, and the next mobile growth lever is turning PUBG from a hit live-service game into a creator-led content platform. World of Wonder has already produced more than 5.4M user-created maps, and planned Naruto and Spider-Man collaborations should bring more players and creators into that ecosystem. If Krafton can pair stronger creation tools with meaningful monetization opportunities, PUBG Mobile can expand its content supply without relying solely on internal teams and licensed events, improving engagement and extending an already unusually long lifecycle.

Playtika’s DTC revenue jumps 63% as Disney Solitaire approaches Bingo Blitz

• Playtika generated $731.1M in Q2 revenue, up 5% YoY, while DTC platform revenue increased 63.1% to $286.9M. Disney Solitaire generated $142.4M during the quarter, up 15.5% sequentially and 288.6% YoY, putting it close to Bingo Blitz’s $145.1M quarterly revenue (down 10% YoY); SuperPlay also became a positive adjusted EBITDA contributor.

• Playtika is now generating almost $287M per quarter through DTC, demonstrating how meaningful webstore migration has become for large mobile publishers. Disney Solitaire is also strengthening Playtika’s results, but its growth has required substantial upfront investment. Playtika front-loaded user-acquisition spend into H1, partly because SuperPlay’s earnout incentivized early investment, and now plans to reduce that by roughly 70% in H2. Management expects Disney Solitaire’s revenue to decline sequentially as a result, but says strong retention should allow the game to remain profitable on the players already acquired. Meanwhile, Playtika’s acquisition terms require up to $1.25B in additional cash payments tied to SuperPlay’s revenue and EBITDA through 2027. A reported sale to Tencent could provide cash and reduce these obligations, but would also surrender Playtika’s strongest growth asset.

Court orders Papaya Gaming to pay Skillz $719M in false-advertising case

• A U.S. court has ordered Papaya Gaming to pay $719M to Skillz Platform after an April jury found Papaya liable for false advertising and unfair business practices. Skillz alleged that Papaya used bots to simulate human opponents while marketing its cash-based mobile games as fair, skill-based competitions. Trial evidence showed that Papaya’s bots outnumbered human players between 2021 and 2024. The court selected the $719M disgorgement of unjust profits instead of the jury’s earlier $420M damages award, since Skillz can recover only once for the same injury. It also denied Papaya’s post-trial challenges and awarded Skillz approximately $10M in legal fees. Papaya says it is considering an appeal, while Skillz cautions that further proceedings could affect when or how much it ultimately collects.

• The judgment is a major warning to cash-prize mobile game operators that opponent matching is not merely a product-design choice. When players are told they are competing against people, undisclosed bots can expose the operator to substantial liability if they affect prize distribution or spending behavior. That makes clear disclosure and auditable matchmaking essential for any studio using simulated players in monetized competitions. The size of the award could also intensify scrutiny of the wider skill-gaming category, where trust is already central to acquisition and retention.

Other News Items

• $30M Fund for Turkish Mobile Game Studios by Drilo Ventures

• Unity adds dedicated engine support for Netflix Games

• Garena is developing Palworld Online, launching this year

• Arknights: Endfield surpasses $125m in six months on mobile

• Sega and Black White Orange partner to expand Angry Birds’ presence in India

Content Worth Consuming

• H1 2026 Gaming Digital Market Index (Sensor Tower): “The cross-platform snapshot shows the pattern clearly. Worldwide mobile IAP finished around $40B, down 2% year over year, while mobile playtime held at 221B hours. Downloads told a different story, falling 12% to 24B.”

• 5 Numbers Hiding in Plain Sight: Sensor Tower’s Ad Monetization Report (Deconstructor of Fun): “I spent half an hour with Sensor Tower’s Lead Analyst, Sam Aune, working through the charts. Five surprising numbers came out of it. Some are buried in the body copy. Some you only get by dividing two of Sensor Tower’s own figures together. Most importantly, none of them are on the summary slide.”

• GTA has made over $100m on mobile – and now Tencent, Garena and more are moving onto Rockstar’s turf (Mobilegamer.biz): “With the GTA 6 launch looming large over the entire games business, we’ve had a look at Rockstar’s own games and the wider ‘GTA-alike’ category on mobile to see which other studios are trying to ride the wave of hype coming our way this November.”

• The industry makes great games, but nobody can find them (gamesindustry.biz): “The workgroup found that the reason why the discoverability problem hasn’t been fixed is because the solution would be expensive, risky, and go against the short-term interests of almost every party involved. And there’s a deeper issue underneath the discoverability problem: the industry may simply be making too much, and too little of it matters.”

• How SayGames uses game analytics to balance difficulty in hybrid puzzle games (pocketgamer.biz): “In one case, we experimented with adjusting the balance of levels without redesigning them. By slightly softening parts of the difficulty curve, the overall pacing of the game improved. Levels felt less frustrating, and the progression through the game became smoother.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s how we can help, spanning all platforms, genres, and regions:

• Strategy Consulting: Market research, corporate strategy, game and economy design, gamification, live-ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above to learn more. If you'd like to discuss how Naavik can support your team, click the box below or send us a note at consulting@naavik.co.

As you may have seen, we are welcoming our third cohort of gaming experts to our Fractional Talent network! We continue to expand our roster of highly accomplished games industry professionals, bringing an even broader range of roles, seniority levels, and geographies.

Naavik’s Fractional Talent service connects clients with pre-vetted experts across a wide range of disciplines, all available for fractional or freelance support. Naavik handles all legal, billing, relationship management, and service quality control responsibilities, allowing clients to quickly access specialized subject matter experts and embed them directly within their team for hands-on support. If your team has such needs, please reach out.

As a reminder, Naavik’s team will be at Gamescom on August 26 and 27. Reach out if you’d like to meet with us or explore how Naavik can support your team.

Sponsor: Create Next Generation Game Experiences with AWS

AWS for Games gives you everything you need to build, run, and grow innovative gaming experiences. Leaders across the gaming industry rely on AWS’s more than 16 years of games expertise to launch and run some of the world’s biggest games, including Fortnite, Roblox, and League of Legends.

AWS and over 300 AWS Games Partners offer the broadest and deepest set of purpose-built services and pre-built solutions for cloud-based game development, content creation, game backend and infrastructure, player acquisition, retention, and monetization. These tools empower game developers to build healthy global player communities, gain player insights, personalize, and monetize player experiences, to create lifelong fans. Around the world, 750 million gamers play games built on AWS each month.

Learn how AWS can help you build, run, and grow innovative gaming experiences.

Learn More

Inside Meshy's $400M Raise for 3D Assets

Written by Francois Courset, Lead Consultant at Naavik

On July 21, 2026, Meshy closed a $400M Series B at a $1.5B valuation, which the company billed as “the largest in 3D Generative AI history.” Beyond the 3D asset generation space, it ranks among the largest funding rounds for an AI startup serving the games industry, with a reported $40M in ARR. Meshy's raise wasn't an isolated event: Tripo, General Intuition, and HiDream have all announced sizable rounds in recent weeks. Together, these deals signal a dramatic acceleration in AI-in-gaming investment.

So what exactly is crystallizing investor conviction around Meshy? What does its trajectory tell us about how close AI-generated 3D assets are to real production-ready workflows? And what does the wider pattern of mega-raises say about where the AI-in-gaming ecosystem is headed next?

Meshy’s Comprehensive Approach to AI-3D Workflows

Before Meshy became one of the most talked-about names in AI-powered 3D content creation, it was Taichi Graphics — a professional graphics software company founded in 2021 by MIT PhD Yuanming "Ethan" Hu. The pivot came in 2023 when the team bet that generative AI would rewrite the economics of 3D asset creation. Meshy now supports Text-to-3D, Image-to-3D, AI texturing, animation, and API workflows serving game studios, 3D printing, and other creative industries.

Meshy runs its roughly 150-person team across its U.S. headquarters and a significant Beijing operation — a cross-border structure reflected in its cap table. Backers include a who's who of top-tier China-focused venture capital: IDG Capital, Matrix Partners China, Monolith Management, Granite Asia, HongShan (formerly Sequoia China), BAI Capital, and Source Code Capital. The company has maintained an explosive growth trajectory: by July 2026, it had crossed 12M registered users and more than 100M models created, with ARR increasing roughly 12x YoY.

Meshy's 3D Asset Production Capabilities

Meshy aims to democratize 3D asset creation, framing itself as a “Canva for 3D.” At its core, Meshy positions itself as an end-to-end 3D asset creation platform. The platform is browser-based, with a built-in 3D viewer that lets users toggle wireframes, inspect polycounts, and preview textures before downloading assets. Meshy currently serves clients primarily across gaming, academia, XR, and 3D printing. Most gaming clients are mobile studios, although the customer base also includes AAA developers such as Square Enix.

Meshy’s functionality can be broken down into five core areas:

• Ideation: On July 29, Meshy revealed its 3D Agent Beta, designed as a conversational tool for concept exploration that lets users brainstorm ideas, compare visual concepts, and generate downloadable 3D models.

• Asset Generation: Meshy can take text or images as input to generate untextured 3D models (meshes); geometry can be previewed and iterated on before the separate texturing pass. It also offers a multi-view mode, allowing users to upload front, side, and back profiles of a character or prop to lock down the exact geometry and prevent the engine from “hallucinating” the unseen angles. Models can then be downloaded for further editing.

• Texture Generation: Meshy lets users create production-quality PBR textures in seconds by describing a surface or uploading a reference image, either on a Meshy-generated asset or on a user-uploaded one. Meshy automatically handles UV projection, seam placement, and texel density for both, addressing the tedious UV-unwrapping process.

• Rigging and Animation: Meshy can then auto-rig characters and objects, supporting both humanoid and quadruped models. The feature works best with characters generated in a T-pose. Once rigged, users can apply motions from a large preset library.

• Mesh Optimization: A tool that lets users retopologize models to a target polycount while maintaining clean topology.

Additional features include extensive integration and export options, as well as an AI Creative Lab built specifically for 3D printing use cases. Meshy stands out by offering a comprehensive workflow that addresses the most time-consuming parts of the 3D asset pipeline, such as blockout generation, preliminary shapes, asset optimization, and UV-unwrapping. This allows teams to parallelize work within what has traditionally been a waterfall production pipeline.

Meshy's current limitations broadly fall into three areas. First, it remains fundamentally an object generator, with no support for scene or environment assembly, including level layout, world-building, terrain generation, or scene composition. Second, its animation capabilities remain limited, relying primarily on a preset animation library, meaning it lacks features such as keyframe animation, facial rigging and blendshapes, cloth or hair simulation, and non-humanoid or mechanical rigs. Finally, the platform does not yet include downstream production tools such as lighting, rendering, particle/VFX systems, or physics simulation.

Acceleration in Capital Deployment

Over the past few months, a wave of major funding has flowed into AI gaming startups, with most of the investment concentrated in foundational model development — world models in particular (highlighted in orange bars above). As we recently analyzed in depth, these technologies have transformative potential that extends well past gaming and justifies the scope of investment.

That said, not all of that investment is flowing toward world models. A closer look reveals that 3D asset production is having its own moment, with Tripo — Meshy’s closest rival — raising $150M and Hyper3D also closing an undisclosed round. A logical next step would be for Meshy to follow Tripo into world models, which could help explain part of its valuation. However, nothing in Meshy's recent communications suggests that is the company's direction.

Meshy’s Hockey-Stick Growth

The key to understanding Meshy's fundraise is scale. The company says the proceeds will be spent on multimodal foundation-model R&D and global expansion — essentially investing to accelerate growth while strengthening its competitive position. Taken together, these investments reflect proven product traction rather than a speculative bet on AI's long-term potential.

Meshy has been on a remarkable growth trajectory over recent months. It reported reaching $15M ARR in November 2025, then $30M by March 2026, and a reported $40M in April 2026, a textbook hockey-stick curve. That growth is supported by 12M registered users, a healthy client acquisition funnel, and an 85% gross margin. While these numbers are company-reported and hence should be taken with a pinch of salt, product traction is undeniable.

Meshy's cross-border positioning is another key strength, providing access to both capital and talent. It has also helped the company attract major China-based clients, such as NetEase Games, FunPlus, and 37 Interactive Entertainment. China-based companies have a more pragmatic approach to pipeline optimization, massive content production requirements for live games, and a more AI-friendly consumer audience — all of which make the market a strong fit for Meshy's offering. Clients featured on Meshy’s website remain predominantly Asia-based, with few Western AAA clients; this suggests that some segments of the industry still have to take the leap, leaving some growth opportunities in the West and across PC and console for Meshy.

As is typical with Series B, client traction is now proven. A report from Room8 earlier this year claimed that its team was already achieving 10–15% faster prop production and an average 30–40% savings on low-poly optimization with 3D AI generation tools. These are strong enough metrics to build a case for adopting such solutions. Human input is still required for hero assets, but tools like Meshy can address much of the lower-hanging fruit, such as ideation, background props, and base meshes.

What’s Next?

Meshy’s core offering sits within the creator-application layer. Looking ahead, it will have to maintain competitiveness in foundation-model R&D while strengthening its defensibility. It will also have to expand its platform by covering more of the production workflow, while its 3D agents initiative will push the needle toward orchestration.

As it accelerates its global expansion, one of Meshy's biggest challenges will be maintaining its broad "Canva-like" appeal while meeting the increasingly complex workflow requirements of AAA game development. The fresh capital will also help accelerate expansion toward adjacent categories such as animation, movies, architecture, or industrial applications. Meshy’s cross-border positioning might also raise regulatory challenges, while training data governance risks could pose regulatory and legal challenges, as illustrated by recent high-profile cases.

From a product innovation perspective, world models could be a natural extension of Meshy’s focus, allowing it to support environment generation features and aligning with its mission to let users “step inside what you imagine.” A more active innovation area lies in Meshy’s recent expansion toward gameplay use cases through interactive multimodal content experiences. At GDC, the company unveiled an AI-enabled Survivor-like game called “Black Box: Infinite Arsenal.” The game allows users to prompt actions in-game and dynamically generate weapon mechanisms and game scripts, producing a randomized, evolving experience each run.

From a broader industry perspective, Meshy signals that AI-assisted asset generation has crossed the commercial usefulness threshold. As one of the more technically demanding AI applications, this suggests that other application areas are already crossing similar thresholds (such as Ad creative production). There are strong reasons to believe that Meshy is the tip of the spear, and similar acceleration will likely follow across other gaming use cases such as audio or 2D assets. We might be witnessing the very first steps of gaming AI’s maturity phase, where the focus shifts from experimentation toward increasingly broad and standardized deployment. We'll continue tracking these developments.

Top News Items

Written by Devin Becker, Consultant at Naavik

Voices for Games Pays Voice Actors for AI Versions of Their Performances

• Voices.com launched Voices for Games, an AI voice platform that takes a fundamentally different approach to the synthetic voice gold rush with every AI voice built on the performance of a real, consenting actor who is licensed and compensated whenever their voice is used. The platform offers two products. The first is Branded AI Voice for Characters, a fully governed system for studios that handles casting, directed recording, exclusive licensing, and production support for everything from cinematic leads to NPCs across 110+ languages. The second is Licensed Character Voice Datasets for model builders, offering 450+ character archetypes and 800+ hours of directed voice data with clean metadata. Supercell already uses Voices for traditional casting and is now working with the AI dataset. Critically, this model aligns with the SAG-AFTRA agreement’s three principles of consent, compensation, and actor control.

• According to Amplified’s 2026 voices report, 79% of decision-makers say AI voices should come from real, attributed talent, but most existing AI voice platforms still rely on scraped or fully synthetic approaches. The gap between what studios want (legally defensible, brand-safe AI voice) and what most providers offer is wide, and Voices is hoping to close it. For games specifically, the implications are significant, with AI-driven NPCs and conversational characters becoming standard, through products such as Epic's Conversations and Latitude's Voyage. The demand for high-quality, ethically sourced voice data will likely scale alongside these features. Realistically, this splits the market rather than reshaping it. Studios with recognizable talent or a brand worth protecting have a concrete reason to pay for a clean pipeline to avoid potential legal action from performers, while smaller and international teams filling out ambient NPC barks will keep buying whatever is cheap and good enough. Voices is selling provenance, which only carries a premium for buyers who need to prove it or want the PR benefits.

Teen Founder Raises $3.2M for AI Game Development Startup ForgeGUI

• ForgeGUI, an AI-powered game development tool built by 18-year-old founder Giulio Greco, raised $3.2 million at a $40 million valuation. Greco, a high school senior with eight years of UGC development experience across Minecraft and Roblox, launched ForgeGUI in March to help individual creators quickly build 3D game assets using models from companies like OpenAI. The platform has already crossed 300k users, and Greco believes it will hit 1M by year’s end. He has exactly one employee besides himself, has moved from Texas to San Francisco, and has no plans for college.

• Gaining 300k users in four months for a tool targeting UGC game creators suggests real demand for AI-assisted 3D asset generation at the individual-creator level, where traditional tools like Blender or Maya remain too complex and time-intensive. This is the same democratization story playing out across the creator ecosystem, from Roblox’s own agentic AI tools to Meshy and Luma’s 3D generators, but ForgeGUI’s growth rate at this stage is notable. As much as traditional game devs push back against AI, the smaller UGC creators embrace it, further democratizing game development. A $40M valuation on $3.2M raised is aggressive, and the company is extremely early with just one engineer. VCs are clearly betting that the winner in AI-for-UGC asset creation captures an enormous market as platforms like Roblox, Fortnite Creative, and their successors continue to grow their creator bases. Whether ForgeGUI specifically can hold that position against well-funded competitors remains an open question.

AI Startup Versus Emerges From Stealth to Fight Streaming Churn With Gamification

• Versus emerged from stealth with a pitch that sits squarely at the intersection of games and streaming. Its AI-powered interactive overlays let viewers “play along” with content they’re watching. When users point their phone at a live sport, series, or film, Versus generates games, trivia, polls, and prediction challenges layered on top of the viewing experience. The company has already worked with Disney+, Paramount+, HBO Max, the NFL, Lionsgate, and Reddit, and claims results including 16% longer session lengths, retention lifts of 4.5% to 9.5%, and content start rates exceeding 53% against a 2% industry average. Notably, CTO Jesper Jenson was previously GM of Yahoo Games and a senior executive at Electronic Arts.

• This matters for gaming because it represents gamification flowing in the opposite direction. Instead of games borrowing narrative techniques from film and TV, streaming platforms are borrowing engagement mechanics from games to solve one of their most expensive challenges: subscriber churn. The economics are compelling, as subscriber reacquisition costs across the streaming industry run into the hundreds of millions annually. Versus is also launching FourPlay.ai, a direct-to-consumer interactive platform focused on live sports and events, which could create an entirely new category of second-screen gaming. For game companies, this reinforces the gamification concept that the engagement patterns they’ve refined for decades (leaderboards, predictions, real-time competition) are becoming the template for retention across all digital entertainment. Gaming companies could find new licensing and partnership opportunities in developing engagement mechanics for adjacent entertainment verticals.

Nvidia Launches Omniverse Libraries and Showcases Graphics Research Breakthroughs at Siggraph

• At Siggraph 2026, Nvidia announced Omniverse Libraries, a collection of disaggregated, agent-friendly software components that give AI agents the tools to build simulation-ready 3D worlds. This is the latest step in a nearly decade-long effort to break Omniverse into increasingly accessible pieces. The new libraries include ovrtx for RTX sensor simulation (camera, lidar, radar), ovphysx for GPU-accelerated physics, and CAD-to-SimReady skills for converting design files into simulation-ready assets via OpenUSD. Partners SideFX and PTC are already integrating these into Houdini and Onshape, and a Blender integration blueprint is available on GitHub. Alongside the platform news, Nvidia presented 21 accepted technical papers, highlighted by MotionBricks, a real-time motion model trained on over 350k clips that can drive both in-game animated characters and a physical Unitree G1 humanoid robot using the same underlying system.

• Nvidia is continuing to try and build an infrastructure layer where game development, robotics simulation, and physical AI all converge. The same tools that let a game studio generate physically accurate environments for a racing game also let an automaker train autonomous vehicles. That convergence is strategic for Nvidia because it means simulation workloads serving a growing robotics market will run on the same stack as the games and content creation market, giving both sides more scale and better tooling. For game developers specifically, the agent-friendly architecture matters most. As AI agents are increasingly used in 3D pipelines (a trend Roblox, Unity, and Epic are all pushing), the studios and tools that adopt simulation-grade standards could produce more robust, reusable content. Nvidia's goal is for OpenUSD and Omniverse to become foundational components of that future.

Other News Items

• D&D rival Chaosium removes AI errors added to new game “without our knowledge”

• EA boss awarded $38.6M after year of franchise success, studio layoffs, and GenAI adoption

• Unity says Unity 7 game engine coming in Q1 2027

• SOCAN partners with Musical AI on music creator attribution, pay and transparency

• LinkedIn actually adds a ‘seems like AI slop’ button

• Adobe’s ‘natural look’ camera app embraces generative AI

• Artists are lawyering up against AI slop, and some are even winning

• Nvidia, Microsoft launch open AI security alliance — without OpenAI, Google, or Anthropic

• Meta is testing an AI bedtime story app for people with no imagination

Content Worth Consuming

• The Most Honest AI Adoption Story in Gaming (Deconstructor of Fun): “Sunrise Village is not an AI growth story. Downloads have halved. Revenue is down by a fifth. But it is a tale of profitability. The team size went from 25 to 3. The game went from deep red to green. And that green is what allows InnoGames to invest in new games.”

• Network expert Glenn Fiedler warns game programmers to fix game code vulnerabilities that AI can now find (Gamesbeat): “I passed Claude (Code) Fable 5 over my network libraries starting Monday last week. Starting at 11 p.m., I was up all night until 6 a.m., and have been working non-stop since then to fix everything up. There is something massive coming. First, AI coding is real and it’s going to expose massive security flaws. in game libraries (like my own) that were 10 years old and I thought were completely secure.”

• How China’s game industry is using AI, from art to ad-buying (Pocket Gamer): “That claim illustrates one of the central tensions surrounding adoption. The technology can allow companies to support larger player communities, but it may also reduce the number of employees required to operate those services. Lyndon says similar capabilities are also becoming available to companies that previously could not afford the kinds of systems used by China’s largest publishers.”

• Liquidnitro Games hires ex-Tencent Games global CTO for AI-enabled game production (Gamesbeat): “Sandeep contacted me and showed me some of the technologies and shared a lot of the thoughts that the team have been putting behind their technology and everything they are working on. I was really blown away. I think you have a lot of AI hype right now, and you have a lot of different things. But what I saw in what Sandeep showed me was really real in terms of taking a live service and really transforming the timelines, compressing the timelines, reducing the operating cost, growing the games, replicating and enhancing what the team had been doing at EA, but really taking it to the next level thanks to the power of AI.”

• AI for Games: KRAFTON Highlights ‘Three Fronts’ (Invenglobal): “We have built games using physics engines and graphic rendering until now, but world models that generate video could potentially replace game engines in the long term. Who knows? We might end up using game engines and world models together to create games.”

• Building an AI game testing agent with Amazon Bedrock (AWS for Games): “We showed how to build an autonomous QA agent on Amazon Bedrock that connects to a running Unity game, reasons about game state, and executes test scenarios without human intervention. We covered connecting to a game’s internal object hierarchy, a perceive-reason-act-reflect loop inspired by industry research, and a framework of game-agnostic tools that separate game knowledge from execution logic.”

• Ethical, low-power, and on-device: Studio Atelico’s vision for the future of generative AI (Gamesindustry.biz): “De Nucci expounded the benefits of on-device AI: namely, low power usage and the lack of a need for constant online connection. Plus, he pointed out that vast amounts of training data simply aren’t needed when you only want the model to do one thing, like character creation.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the games industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects spanning market research, corporate strategy, game and economy design, gamification, live-ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can support your team, click the box below or send us a note at consulting@naavik.co.

Today, we are launching our new biweekly newsletter focused on mobile F2P gaming! It will become Naavik’s primary hub for exploring the games, genres, companies, and trends shaping the future of mobile play.

A few quick details: This newsletter segment will be published via Substack under our Naavik publication hub, and it will come out every alternate Tuesday, between our AI x Gaming issues. We also plan to port Naavik Digest to Substack soon so that we can have all our newsletter content in one place. Stay tuned for that.

Lastly, we recognize that not everyone is interested in every aspect or platform of the gaming industry. If mobile F2P isn't for you, we’ve got you covered! Substack makes it easy to adjust which newsletter segments you receive. Here's how: go to Settings, click Naavik, and toggle which notifications you receive from our Substack.

Enough preamble. In today's issue, we cover the biggest developments in mobile gaming over H1 2026. Plus, make sure to check out our latest interview, which dives into Subway Surfers' longevity with SYBO CEO Mathias Gredal Nørvig.

H1 2026: Mobile Gaming’s Stability Illusion

Written by Harshal Karvande (Principal Designer at Rovio) and Abhimanyu Kumar (Co-founder of Naavik)

Mobile gaming entered 2026 with a paradox. According to Sensor Tower data, global downloads fell to ~22B in H1 2026, a significant -3B YoY (-12% YoY) reduction in volume. On the other hand, revenue hit ~$28B, a relatively modest -$0.6B YoY (-2% YoY) decline in value. Specifically in Q2 2026, downloads fell to pre-IDFA and even pre-pandemic levels to ~10.8B (-12% YoY), while revenue was slightly less impacted, dropping to ~$13.3B (-5% YoY). The market is clearly losing downloads faster than it is losing revenue, which makes H1 2026’s revenue story look stable even though some of its foundations seem to be shifting underneath.

To understand what’s really happening, this article will dig deeper than the aggregate numbers above — analyzing movements across platforms, genres, subgenres, and games. Note that the data does not include DTC, China Android, or ad revenues. Also, the market size numbers come from Sensor Tower’s “Market Size” feature, which showcases a subset of the entire mobile gaming market. While the aggregate numbers look smaller than mobile gaming’s true market size, we’ve confirmed that analyzing this subset is representative of broader market trends.

Let’s dive in.

Downloads: -13% YoY, Android-driven

When splitting H1 2026’s download decline by platform, Android was the major culprit, declining at a -13% YoY rate and contributing to ~92% (~2.8B) of the entire decline. iOS drove the rest (~8%, ~0.2B, -6% YoY).

The geographic breakdown makes this concentration even clearer. Tier 2 and 3 markets accounted for ~88% of Android’s decline and ~80% of the combined decline across platforms. The top 5 contributing countries were India, Brazil, Indonesia, Mexico, and Turkey.

Given the above, it is not surprising to see that the typical high-volume genres that receive installs from Tier 2 and 3 countries were primarily responsible for Android’s H1 2026 download decline. The Arcade genre accounted for ~30% of the drop, followed by Simulation, which drove an additional ~22%. Lifestyle, Action, and Tabletop also dropped. Collectively, these five genres represented ~78% of Android’s decline.

Digging deeper, Android’s genre-specific declines were concentrated within certain subgenres that typically garner high volumes of Tier 2 and 3 region installs. For example, Arcade’s decline was driven by Platformer / Runners (~46% share), Other Arcade (~20% share), and Mini-game Hub (~12% share). Similarly, Simulation’s decline was mostly driven by Simulator (~38% share), Driving / Flight Simulator (~23% share), and Sandbox (~15% share).

That said, it should be noted that across most of these subgenres, the download decline was subgenre-wide (spread across multiple titles). This was primarily due to two simultaneous effects:

1. In certain subgenres, many short-lived game launches faded faster than newer releases could replace them. For example, Other Arcade, Mini-game Hub, Simulator, and Driving / Flight Simulator all followed this behavior, with “pop-and-drop” games contributing 50-60%+ of the subgenres’ decline.

2. In other subgenres, their weakness was primarily caused by older games fading. For example, in Platformer / Runners, mature titles drove ~85% of the subgenre’s download decline, while rapid decay in more recent launches drove the rest.

Overall, H1 2026’s 12% YoY download decline can be summarized as follows. Android experienced a much sharper contraction than iOS, generating ~92% of the combined platform decline. The correction was concentrated in high-volume Tier 2 and 3 countries and a small set of genres and subgenres that usually receive a large share of their installs from these territories. Within the largest declining subgenres, a mix of short-lived launches and aging titles saw their download volume fade faster than newer releases could replace them.

That said, what’s very striking to see in H1 2026’s results is that all genre downloads declined YoY at a combined platform level. Unfortunately, this seems to be part of a longer-term structural shift in mobile gaming’s top of the funnel that goes beyond post-COVID and post-ATT effects. It’s not only concentrated to Tier 2 and 3 markets but also seems to impact Tier 1 regions. We will explore a few hypotheses behind why this might be happening (and what it means for mobile gaming’s future) in the final section of the article.

Revenue: -2% YoY, But Not a Healthy Market

Given that H1 2026’s IAP revenue was relatively flat (-2% YoY) alongside downloads falling -12% YoY, the question becomes: What helped make up the difference?

The difference came from revenue per download, which rose +11% YoY and balanced out the significant downloads drop to result in relatively flat YoY revenue. In other words, applying H1 2025’s revenue per download to H1 2026’s downloads volume would’ve reduced its revenue by ~$3.3B. However, H1 2026’s increased revenue per download restored ~$2.8B of that, leaving the -$558M (-2% YoY) decline.

Yet, higher revenue per download math masks significant genre movement underneath. Some of mobile’s largest revenue-generating genres, like RPG, Casino, Strategy, and Action, lost a combined ~$1.6B YoY, while Puzzle and Shooter offset ~$1.3B of those losses. A closer look at those genre movements reveals several key observations.

Note that DTC revenue outflows, which are not tracked in Sensor Tower’s data yet, might also be contributing a bit to the genre IAP revenue movement discussed below. According to Appcharge, 2025’s global mobile DTC revenues were estimated at ~$17B, which accounts for ~15% of the $113.3 billion mobile gaming IAP market in the same year. And some subgenres are adopting DTC payment channels more aggressively than others — for example, RPGs (across multiple subgenres), Casino games (especially Slots), and Strategy titles (especially 4X Strategy ones). However, it likely doesn’t change the broader trends and takeaways noted through this analysis.

#1 — RPGs’ decline was severe and genre-wide: RPGs’ -$664M YoY decline fully explains mobile gaming’s H1 2026 revenue decline and was driven by multiple key revenue driving RPG subgenres:

• MMORPGs dropped by -$455M YoY (~69% of total genre decline), due to the portfolios of established Asian publishers seeing natural aging over time. For example, NC Corporation’s (formerly known as NCSoft) Lineage M, Lineage W, and Lineage 2M contributed to ~30% of RPGs’ decline, while other mature Korean MMORPGs also declined significantly. This is not uncommon for MMORPGs, but new major releases are necessary to make up for the gap to ensure overall subgenre stability.

• Turn-based RPGs dropped by -$266M YoY (~40% of total genre decline), primarily due to weakness in various mature titles like Honkai: Star Rail (-$69M), Fate/Grand Order (-$40M), and RAID: Shadow Legends (-$37M). Across the three titles, their declines were driven by either MAU erosion (natural to mature titles), underperforming ARPDAU (weaker live-ops roadmap), or a mix.

• Squad RPGs dropped by -$157M YoY (~24% of total genre decline), again due to declining established titles like CookieRun: Kingdom (MAU erosion), DC: Dark Legion (post-launch year decline), and Hero Wars (MAU erosion).

• Open World Adventure offset some portion of the above with a +$173M YoY bump, mainly due to Arknights: Endfield’s and Tencent’s Roco Kingdom’s strong launch performance in China.

#2 — Casino’s decline was Slots and Monopoly Go! driven: Casino saw a -$376M YoY decline in H1 2026, with two subgenres accounting for ~95% of the genre's decline:

• Slots was down by -$233M YoY, contributing to ~62% of Casino’s decline. It was spread across numerous incumbent games, such as Slotomania, House of Fun, DoubleDown Casino, Huuuge Casino, and more. While all these games face growing maturity-related declines (and DTC revenue outflows contributing a bit to the numbers too), most Slots game publishers also lost the ability to retarget lapsed payer cohorts post-ATT to help replenish churned cohorts and thereby keep the subgenre afloat. However, some Slots game publishers were able to navigate this UA hurdle better than others — see our interview with SciPlay.

• Coin Looters was down by -$125M YoY, contributing to an additional ~33% of Casino’s decline. Monopoly Go!’s slow downloads, DAU, and ARPDAU deterioration, likely due to a rapidly scaled game gradually maturing and normalizing, explained all of it.

#3 — Strategy suffered a Pokémon problem: Strategy experienced a -$319M YoY decline with the following taking place:

• Pokémon TCG Pocket fell by -$300M YoY, which accounted for ~94% of the genre’s decline. The game dropped sharply from its launch-period highs as downloads, DAU, and ARPDAU contracted during its second year of live-ops.

• 4X Strategy was up by +$193M YoY, but beneath that were approximately +$974M of gains and -$781M of declines. Kingshot and Last Z: Survival alone generated roughly +$576M of the gains. However, mature leaders such as Last War: Survival and Whiteout Survival (!) subsequently absorbed much of that growth.

#4 — Action witnessed four subgenre corrections: Action saw a -$250M YoY decline with four subgenres (and key titles within them) driving most of it:

• Shoot ‘em Up was down -$101M YoY (~40% genre decline contribution) with ~80% of it explained by weak performance from Habby’s Achero 2, Survivor.io, and Archero. This is not surprising since even the best hybridcasual games today haven’t yet figured out how to fully evolve out of hypercasual‘s typical UA-fueled front-loaded performance, and scale revenue beyond the first three to six launch months. Unfortunately, Habby’s portfolio performance also defines the subgenre trends due to its dominating revenue market share.

• Dungeon & Fighter Mobile/Origin lost around -$97M YoY, more than the Beat ’em Up subgenre’s net decline. Almost all of this came from the game’s post-China launch year decline.

• Zenless Zone Zero lost -$40M YoY, while Solo Leveling: ARISE lost -$18M YoY. Those two franchises explain most of Hack and Slash’s decline before new Bleach releases provided offsets. Hack and Slash‘s revenue performance has always been major launch-driven, as the subgenre’s stagnating audience volume buys into new-release hype before cycling through titles quickly.

• NetEase’s Identity V versions lost -$40M YoY due to a weaker live-ops roadmap, explaining Asymmetric Battle’s decline.

#5 — Puzzle’s rise was mobile’s MVP in H1 2026: Puzzle added +$915M YoY, with two subgenres accounting for nearly all of that growth (as we also previously covered here):

• Merge 2 was up +$640M YoY, primarily driven by the UA fuel behind Microfun’s Gossip Harbor (+$320M) and Century Games’ Tasty Travels (+$132M).

• Sort was up +$303M YoY, mainly driven by a newer cohort of games, such as Pixel Flow (+$125M), Magic Sort (+$33M), Yarn Loop (+$22M), Sand Loop (+$18M), Marble Sort (+$17M), and Gecko Out (+$16M). Most of them, except for Magic Sort, are relatively recent launches and continue to scale with heavy UA.

• Unfortunately, all other Puzzle subgenres collectively declined by ~$28M, including staples like Match Swap, Hidden Objects, and Merge 3.

#6 — Shooters transformed due to Tencent’s new China-release cohort: Shooters grew +$218M YoY, almost entirely because FPS / 3PS increased by +$225M YoY. This was driven by three Tencent-operated releases — Delta Force, Valorant Mobile, and Reverse War: The Future. Over H1 2026, Delta Force significantly grew in China, while Valorant Mobile and Reverse War brought new revenue to the genre as both launched after H1 2025. However, that single-publisher, single-geography growth hid large incumbent declines, including CrossFire Mobile and Call of Duty: Mobile.

Overall, H1 2026’s relatively flat -2% YoY revenue performance shouldn’t be interpreted as a sign of mobile gaming successfully extracting more value out of a dropping downloads base. Puzzle’s outperformance and the Shooter revenue bump masks a common thread across the various declining genres discussed above — growing incumbent weakness and new releases not always being able to make up for the gap in a sustainable long-term fashion. This is most notably seen in the RPG genre, which entered freefall from 2022 onwards as its incumbents weakened and new releases were unable to take their place. We’ll explore this common thread (and why it might be occurring) in the next and final section of the article.

What It Means for Mobile Gaming’s Future

In our eyes, H1 2026’s performance showcases the symptoms of two fundamental problems the mobile gaming industry is currently facing.

First, H1 2026 saw a universal top-of-the-funnel decline across all genres, and it is unfortunately the continuation of a multi-year structural decline in install volume across platforms, with Android being the key contributor. The image below illustrates two downloads market phases:

1. 2020-2023 Plateau: Annual downloads plateaued between ~55-56B, post-pandemic normalization. However, while iOS faced a double whammy post-pandemic and post-ATT download decline, Android’s Tier 2 and 3 markets grew enough to keep the combined platform view broadly flat.

2. 2023-2026 Structural Decline: While iOS downloads continued to decline, Android’s Tier 2 and 3 download engine also began to weaken, after peaking in late 2023. Rolling-four-quarter downloads fell from ~56B in Q3 2023 to ~46B in Q2 2026, a -17% contraction, ~94% of which was Android-driven. This essentially made Android the center of gravity for mobile gaming’s downloads downturn, with weakness in Tier 2 and 3 downloads driving ~80% of it.

Given that Tier 2 and 3 inorganic install share remained relatively flat during a period of significant downloads loss, a mix of two effects could be accelerating this volume decline:

• External factors like the attention wars could be driving low mobile gaming intent install volume to other forms of entertainment — an effect that would even extend to high-value Tier 1 market installs.

• Inorganic install spending pullbacks could be occurring due to rising CPIs, which in turn create a poor ROAS environment. According to Sensor Tower’s Gaming Digital Market Index report, over H1 2026, gaming digital ad spend was up +8% YoY (~$7B), and ad impressions were up +14% (~1.4T). Taking those numbers within the context of ad network market share increasingly consolidating and download volumes dropping would indicate CPIs rising at a faster rate than product LTVs can improve to maintain healthy ROAS results across key downloads (Tier 2 and 3) and revenue (Tier 1) markets.

Second, H1 2026’s revenue results tend to showcase a sobering reality about mobile gaming’s mature state:

• The market has become increasingly dependent on an aging base of long-standing incumbents. Games five years or older now generate ~57% of the revenue among the top 200 grossing titles — roughly 70% of total mobile gaming revenue. That share has risen rapidly over the years, underscoring the aforementioned market dependence.

• However, the above doesn’t mean the revenue impact of younger hit games has disappeared. In H1 2026, games less than two years old added roughly $2.3B YoY, replacing 99% of the roughly $2.4B lost by incumbents five years or older. The bigger concern is how rarely successful new launches mature into enduring incumbents. Roughly one in four young top 200 grossing entrants remain on that list long-term and eventually become an incumbent five years or older.

Simply put, while mobile gaming can still generate commercially meaningful launches, few become the durable next-generation incumbents. This raises the dependence of market health on existing incumbents, while increasing the incumbent replacement burden for new games.

All the above reminds us of Supercell CEO’s words from his recent annual letter: “It’s an industry coasting… getting very good at optimizing what already exists… but here’s the reality: live game excellence alone doesn’t grow an industry. It maintains one. For the market to truly expand, we need to bring new players in… And that requires innovation. New genres. New ways to play… That’s what Clash Royale did in 2016. That’s what Pokémon GO did that same summer. The industry needs more of those moments.”

His annual letter is titled “The Best Games Haven’t Been Made Yet”. While we can understand a mobile gaming entrepreneur wanting to hold an optimistic stance, this analysis makes us think about the opposite — have the best games already been made? That’s, of course, a very pessimistic view for an industry that rests on creative foundations that inherently have no design ceiling. However, mobile gamers (new or old) have one important ceiling — dedicatable daily time to mobile gaming. That means they have to make two choices: whether to spend their limited time on mobile games instead of other forms of entertainment, and, if so, which games earn that time. If those games remain long-time incumbents, then natural churn eventually takes over — and there is a chance the mobile gaming industry loses those players entirely, unless a new game takes the incumbent’s place and wins the lost players’ time back.

In other words, Ilkka is broadly right, even though what he’s rallying the industry towards is one of the hardest things to achieve in gaming. At the same time, it doesn’t look like the alternative path is getting mobile gaming anywhere healthy either. So current and future mobile game developers who care about returning mobile gaming to healthy growth (and creating highly durable long-term mobile gaming businesses while at it) may as well be ambitious and solve the harder problem. Mobile gamers are likely eager to experience another market-defining game like Clash Royale for the first time too.

Top News Items

Written by Max Abrahamsen, Consultant at Naavik

3rd Party App Stores Can Launch Inside Google Play

• Google and Epic Games have withdrawn their request to replace a U.S. court injunction with a negotiated settlement, clearing the way for rival Android app stores to be distributed through Google Play from July 22. The injunction follows Epic’s antitrust victory, which found that Google had maintained an illegal monopoly over Android app distribution. Google is also opening access to its Play catalog, meaning eligible third-party stores can list apps and games already available on Google Play unless developers opt out. Participating stores must pay a $5,000 annual review fee, operate only in the US, remain open to eligible developers, publish trust and safety policies, and keep malware install attempts below 1%.

• The ruling creates a materially easier route for alternative game stores to reach U.S. Android users. For mobile game companies, distribution through Google Play is the most important part of the ruling, because it gives rival stores access where users already search for apps. That could make stores from Epic and other publishers easier to install, while creating more competition over fees, discovery, payments, and developer terms. The withdrawal also preserves developers’ right to link users from Play-distributed apps to external downloads. However, it does not settle whether Google can charge fees on those link-outs, leaving a potentially important DTC economics dispute unresolved. Broad adoption of alternative stores still looks unlikely since most consumers have little reason to leave Google Play. The more probable outcome is a small number of publisher-led stores building meaningful but niche audiences around exclusive content rather than a fragmented mass market.

Tencent in Talks to Acquire Playtika's SuperPlay in Deal Worth up to $1.5B

• Tencent is currently in negotiations to acquire mobile game developer SuperPlay from Playtika in a deal valued between $1B and $1.5B. Playtika originally purchased the studio behind Dice Dreams and Domino Dreams in November 2024 for $690M in cash. That initial deal included a performance-based earnout structure worth up to $1.25B tied to the studio’s financial targets between 2025 and 2027. SuperPlay subsequently launched the highly successful Disney Solitaire, which helped the studio generate $573M in revenue in 2025, beating the baseline earnout target by 67 percent. Under the proposed terms, Tencent would assume responsibility for the ongoing earnout payments to SuperPlay’s founders, removing a major financial liability from Playtika’s balance sheet.

• This transaction highlights a rare scenario where a subsidiary’s rapid success has become a financial burden for its parent company. Playtika is facing $2.3B in debt maturities over the next few years and has suspended its dividend to preserve cash. Selling one of its most valuable assets marks a significant strategic reversal, as the SuperPlay acquisition was originally intended to diversify Playtika away from its declining social casino titles and into the growing casual games market. For Tencent, the deal represents an opportunity to strengthen its portfolio outside of midcore and strategy genres, capturing a proven developer in the casual space. Regardless of the transaction outcome, the deal shows how heavy debt can reduce a company’s flexibility, even when it owns valuable assets.

Vgames Launches $500M Growth Financing Platform for Games Companies

• Israeli gaming fund Vgames has secured $500M, backed by Phoenix, to launch a growth financing platform for gaming and consumer companies. The model is designed for businesses that have moved beyond product development and need capital to fund user acquisition and international expansion without selling more equity. Instead of following a fixed repayment schedule, repayments are tied to revenue generated by the user cohorts financed through the platform. Vgames, which has invested in more than 50 companies since 2020, is working with General Catalyst to identify and assess potential recipients and plans to provide more than $500M in additional financing over the coming years.

• Revenue-linked financing gives established studios another way to scale profitable campaigns while preserving ownership, although it still depends on reliable forecasting and disciplined unit economics. Vgames’ move also reflects a broader shift toward specialized financing products for mature games businesses that need growth capital rather than another traditional venture round. Keep in mind that traditional venture funding in video games has plummeted as investors have found better-fitting (and more risk-controlled) ways to invest in games rather than the equity of the companies that make them. Although we expect game financing to continue growing, increasing competition for top-tier opportunities will likely compress margins, push funds beyond gaming, and gradually loosen underwriting standards.

Unity 7 Is Coming Q1 2027

• Unity will release Unity 7 in Q1 2027, following an early beta in December 2026. Unity promises that the new engine will serve as a direct continuation of Unity 6, ensuring no project rebuilds, new programming languages, or breaking transitions. Built on a modernized CoreCLR foundation, it will deliver near-instant Play Mode, reload only modified code, and decrease shader build times by up to 90%. Unity 7 will also introduce real-time global illumination that scales from high-end PCs to mobile, alongside a new command-line interface, public API, and free MCP server that lets external tools and coding agents interact more directly with Unity projects.

• For mobile developers, Unity 7 combines production improvements with a more integrated growth stack. Native direct-to-consumer purchases, no-code webshops, and unified product catalogs will feed transaction data into Vector, Unity’s AI-driven advertising system. This could make it easier for studios to operate web stores and use first-party purchase data to improve user acquisition, while reducing dependence on platform payments. The faster Editor and agent integrations may also shorten iteration cycles for smaller teams, but much will depend on whether Unity delivers these features reliably and maintains compatibility across existing projects. After several years of product and pricing disruption, a smooth upgrade may matter as much as the technical improvements.

EU fines Google $1B over Play Store restrictions and search practices

• The European Commission has fined Google €890M for two breaches of the Digital Markets Act. A €430M penalty covers Play Store rules that restricted developers from promoting cheaper purchases on websites or alternative app stores, while a separate €460M fine addresses Google’s favoring its own shopping, hotel, transport, and sports services in Search. The Commission found that Google’s steering fees were too high and lasted too long to comply with the DMA, although platforms may charge for helping developers acquire new customers. Google has been ordered to end both breaches and has 60 days to comply before risking recurring penalties of up to 5% of its worldwide turnover.

• For mobile game publishers, the important outcome is the EU finding that Google restricted app developers from promoting cheaper offers and completing transactions through web stores or alternative app stores. This directly affects mobile gaming publishers because games generate substantial in-app spending and increasingly use direct-to-consumer stores to reduce platform fees and own the customer relationship. The likely outcome is a narrower fee structure that makes webstore purchases more attractive, although Google will probably continue seeking compensation for users acquired through Play.

Other News Items

• Roblox Brings Prompt-Based Game Creation to Mobile

• Ukie welcomes UK government’s IP-backed lending plans for games studios

• MTG reports strong Q2 results driven by D2C growth

• Wemade controlling stake sold for $593m

• Pokémon Go revenue skyrockets towards a record high during Go Fest 2026: Global

• Royal Kingdom surpasses $750m with 42% of all revenue made in H1 2026

• Netflix kids’ mobile engagement rises 600%

Content Worth Consuming

• India’s State of Play (Naavik, MIXI): “India has emerged as one of the world’s most compelling gaming markets. Its gaming market is maturing, with mobile gaming alone (excluding Real Money Gaming) generating $1.1B through 2025 and projected to reach $2.4B by 2029. India’s long-established scale is now being matched by rising player spending and a widening mix of genres”

• Gaming Deep Dive: Ad Monetization Report (Sensor Tower): “Navigating this market requires understanding its uneven distribution. Mobile ad revenue is highly concentrated across a few specific genres, networks, and product strategies. Puzzle games remain the primary driver of the ad economy, and your choice of ad network can dictate your genre’s access to the right audience. For hybrid-casual titles, deciding to prioritize ads versus in-app purchases (IAP) is a fundamental business decision that defines your ultimate revenue potential.”

• The Winning Portfolio Strategies of 2026 With Two and a Half Gamers (PocketGamer.biz): “At PGC Barcelona, Jakub Remiar broke down the portfolio strategies quietly winning the casual market, from templatisation to knowing exactly when to copy.”

• Google on its new Play Store fees and that new Level Up program (Mobilegamer.biz): “As we’ve said before, the policy changes aren’t the decisive victory Epic claims they are, but they still represent progress. Developers can now claw back a little more of what was a flat 30% IAP cut if they use another payment provider, and from players making installs from now onwards. Devs can get further discounts if they join Google’s Level Up program, though that requires developers to jump through a new set of compliance and technical hoops.”

• Learn all about China’s games industry in our free region report (Pocketgamer.biz): “The report is free to download with insights from industry experts on the state of the sector, data exploring consumer spending in the mobile games space, as well as our newly unveiled Top 30 China Game Makers 2026 list.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s how we can help, spanning all platforms, genres, and regions:

• Strategy Consulting: Market research, corporate strategy, game and economy design, gamification, live-ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above to learn more. If you'd like to discuss how Naavik can support your team, click the box below or send us a note at consulting@naavik.co.

Before we jump into this week’s issue, we’re excited to welcome Sett as a partner! Sett is an agentic AI platform for mobile game marketing that creates and optimizes user acquisition creatives, helping mobile game teams find winning creatives faster and grow their games into global hits. To learn more, check out their podcast Unpredictable Hits.

Also, the Naavik team will attend Gamescom next month on August 26 and 27. Reach out if you’d like to meet, discuss the AI ecosystem, or explore consulting support from Naavik.

Sponsor: How Agentic AI Drives Efficiency in Game Development

Industry leaders from Xbox, Sega, PlayStation Studios, and EPAM recently gathered to examine what happens when AI stops following scripts and starts shaping experiences. The conversation covered autonomous NPCs, adaptive world-building, AI safety, and the critical role that human creativity still plays when technology handles repetitive work.

Key themes explored:

• From copilot to agent. AI has evolved from coding assistants to end-to-end agentic systems capable of autonomously delivering entire features.

• Creativity amplified, not replaced. AI accelerates prototyping, while taste and craft remain irreplaceable — tools alone never reach a creator’s standard.

• Living, infinite worlds. Games can now adapt in real time to a player’s personality, mood, and history — moving from static narratives to endlessly evolving content.

• Safety first. Player-facing generative AI must be load-tested like live services; poor personalization is “toxic to the experience.”

• New genres emerging. Agentic AI is enabling entirely new AI-native experiences while also reviving text-based adventures, giving player deeper agency through more dynamic interactions.

Read the full report here: Agentic AI in Gaming: Evolving Game Development and Player Experiences | EPAM

Read the Report

The State of AI for Game QA

Written by Francois Courset, Lead Consultant at Naavik

From the moment generative AI entered gaming, QA has been viewed as one of the areas most ripe for disruption. Game professionals rank it among the most exciting AI trends, and 94% of developers believe AI will play a meaningful role in QA’s future, unlocking faster iteration and shorter dev cycles. We’re now seeing bold claims, from EA saying AI now shoulders 85% of its QA testing to Square Enix aiming to automate 70% of QA and debugging by 2027.

So, as usual, we wanted to parse through the noise and paint a full picture of AI for game QA today. Why is the need pressing? How mature is the technology? What will the next unlocks mean for the discipline?

The Case for AI-enabled QA Solutions

Game QA is as relevant as ever. Studios allocate up to 10% of a game’s total budget to QA. And because testing needs swing wildly across the production timeline, QA has historically been heavily outsourced, turning Game Testing Services into a billion-dollar segment, growing at a healthy ~10% CAGR. High-profile games that faced backlash — like Mindseye, Borderlands 4, or Monster Hunter Wilds — are a reminder of how consequential proper QA execution still is.

Game QA is facing a math problem. Testing scales linearly, but games are exponentially more complex — open worlds, branching narrative, complex physics, multiplayer networking, live-service support, and a growing stack of platforms. 77% of developers admit their last release shipped with less QA than it deserved, and 50% of developers don’t think that QA budgets are growing fast enough to keep pace with growing QA requirements. More efficient QA workflows enable studios to test more of a game, leading to more polished releases, reducing the risk of backlash, and ultimately protecting sales. They also make QA faster and less expensive, addressing one of game development's largest cost centers. That’s the reasoning behind AI-enabled QA workflows: infrastructure that scales without linearly scaling cost, while running around the clock.

Current Maturity of AI for Game QA

The emergence of black-box testing

Automated QA is not new — scripting, bots, and rule-based test harnesses have been standard tooling for years. Ninety-four percent of developers already use some form of it. The challenge is not whether something can be automated, but rather how it can work cost-effectively and consistently.

Game QA process disruption has been slower than other verticals due to a highly bespoke nature tied to the game’s codebase, and a multiplicity of cases making one-size-fits-all approaches challenging. This has resulted in limited reusability across projects and complex integration processes within QA teams. As Christoffer Holmgård, Co-founder and CTO at Modl.ai, explained in an interview with Naavik: “From a maturity perspective, with very deep integration into the title, you can accomplish a lot for an individual game. But the price to pay is in maintaining this whole test harness. It is like having an extra project on top of building your game.”

This led QA solution providers to prioritize black-box testing. Rather than hard-coding every rule, developers can now deploy AI agents that engage with a game, reproduce human behavior, explore edge cases, and catch bugs. These agents can be run in the cloud and employ computer vision to recognize and interact with game elements. Holmgård added: “Having a solution able to cover any game and just play it is a harder problem in many ways. Black-box testing is the approach Modl.ai and most products are taking. It is completely integrationless, with no code going into the game.”

AI Use Cases for Game QA

AI for game QA can be seen as a spectrum of tools, some requiring deep game-level integration and black-box approaches. Key testing technologies include:

• Scripted Automation: Deterministic code that performs prescriptive actions in a limited scope

• Symbolic AI: Explicit rules, logic, and structured models to reason about a game’s behavior and decide whether it matches expectations

• Large Language Models: Models interpreting natural language requests and producing outputs based on these requests

• Reinforcement Learning: Driving behavior by training a machine learning model through trial and error, and goal setting

• Imitation Learning: Supervised reinforcement learning, where a model learns by using inputs from users simply playing the game

• Computer Vision: A system using what’s shown on the screen to derive game behavior

The technical solution alone is not sufficient and needs to be paired with a comprehensive internal infrastructure. As explained by Modl.ai’s CTO: “You have to think through different layers: First is the interface/integration layer which ensures we can interact with all parts of the game. Second is the behavioral layer, which ensures we do actions that are valuable for the QA team. Lastly is the verification layer that checks QA process outcomes and ensures actionable data is collected about the game build.” AI for game QA needs to be approached as a complete overhaul of existing QA processes to establish proper governance, data collection, and feedback loops. This change management angle is key, as only 18% of developers feel their studios are fully prepared to implement AI tools.

The tipping point for AI-enabled QA

Given the focus on cost-efficient and versatile black-box testing approaches, the QA stack currently isn’t entirely addressable. However, significant efficiencies can already be achieved. As Jan Schnyder, co-founder of Nunu.ai, shared with Naavik, “AI for QA is around a year behind other kinds of AI tools. We are approaching the 30% to 50% mark of automatable test cases. And these systems are getting better every month. We are now addressing 50% of the QA stack when at the start of the year it was around 30%.” Automation potential also varies across games: AAA 3D open-world environments remain harder to tackle compared to 2D mobile experiences.

The lowest-hanging fruit currently includes the most repetitive, pattern-based, low-judgment tasks such as test-case generation, bug triage, log analysis, regression checks, etc.

The foundations are in place for broad QA adoption, with use case sophistication expanding by the month. Studios can already achieve meaningful results across some of the easier tasks. The core challenge remains studio readiness to adopt.

Mapping the AI for Game QA Ecosystem

The AI for QA category is early, fragmented, and still waiting for its breakout moment. Startups in the space have raised a combined $35.8M to date, led by Modl.ai and Nunu.ai, but have yet to reach Series B scale. Rare incumbents such as Tencent’s WeTest have the scale to build AI QA infrastructure in-house. Meanwhile, engine makers are exploring the space with Unity Test Runner or Unreal Gauntlet, hinting at a future where they acquire their way into AI QA. While Keywords has some internal AI QA capability, the broader pattern is service providers leaning on startup-built SaaS tools to enhance their offerings, rather than building from the ground up.

Proper case studies are still hard to find, but all point to real cost savings, speed gains, and minimal setup. Nunu.ai got Stormforge to a 30% reduction in QA costs and 5x faster testing in a five-hour implementation. Nodemori is absorbing 40% of QA workload for clients and cutting build validation time in half. Schnyder added: “The biggest value add for our clients is speed, by running their entire QA plans in hours instead of days, which allows them to do much faster iteration cycles while also lowering costs by up to 30%.”

Future Outlook

Looking ahead, AI for game QA is poised to benefit from multiple advances around perception, efficiency, and actions. As Christoffer Holmgård explains: “The next biggest unlock is around perception: Giving models the ability to accurately perceive what’s happening, locating objects and things that are going on in the game in a timely manner. For such use cases, you need an always-on system, which means you are entering computer vision robotics territory, and we are seeing month by month new models becoming available.” From an efficiency standpoint: “AI for QA systems are always on; as such, the costs of running are not neutral. Managing what kind of models to use for what part of the work is becoming key, especially as better models become available at cheaper rates or are able to run on smaller hardware.”

We drafted the emerging potential of World Models recently, and AI for game QA could be a key proving ground. As Schnyder put it: “World Models is the single biggest thing I’m excited about, which will address one of the biggest bottlenecks right now for game QA — the ability to do 3D real-time environments through new architectures that allow for much faster inference speed.”

AI for QA is also bound to have broad-ranging impacts inside and outside games. First, solving for game QA means solving one of the most complex software use cases; learnings and tools can then be applied and expanded to any sectors beyond games. Then, these tools will be more deeply integrated with the rest of a game’s production pipeline. As Christoffer Holmgård pointed out: “I’m also excited to see who cracks the workflow problem. Once we have an issue identified in a game, AI for QA shouldn’t stop there. It should then take action proactively to fix the issue and address it and help the developer.” Finally, these capabilities can also be extended to other areas of game production. Schnyder added: “What we’re building is an AI agent that can play your game and can perform tasks with it. We actually see customers adopting this across different sectors of a studio such as competitor analysis, design document generation, or updating internal knowledge bases.”

Editor’s note: We’re building an AI-powered, human-verified competitive intelligence platform that combines automated live gameplay with market and product data to reveal how your competitor games work and evolve. We’re currently in alpha and focused on serving mobile F2P teams. If this could help strengthen your product / competitive research function, we’d love to meet at Gamescom and showcase our prototype. Request a meeting here.

AI for game QA offers significant cost and efficiency gains, and with new black-box solutions paired with upcoming technical unlocks, it is nearing full maturity. As Christoffer Holmgård puts it: “We’re not at the point where the AI for QA can yet tell you whether a game is good. And that’s probably fine, because that’s what the human test team needs to figure out and communicate. But in the next year or so we’re going to see a lot of movement on the fundamental question of ‘does it work?’”

Partner Spotlight: Unpredictable Hits: The OGs of Mobile Gaming, Unfiltered.

Unpredictable Hits is the podcast where Amit Carmi, Co-founder and CEO of Sett, sits down with the OGs of mobile gaming for the conversations they usually keep off the record. Real talk on the new world of UA and AI, and where the industry is really heading.

Each season is filmed in a different capital of mobile gaming: London, then Barcelona, and now Tel Aviv, home to some of the biggest games in the world and the teams behind them. Season 3 kicks off with Matt Veysberg, Director of Global Creative Marketing at Plarium.

Previous episodes have featured Jon Bellamy (CEO, Jagex), Akin Babayigit (co-founder, Tripledot), Yair Melamed (SVP Games, Zynga), Zeynep Tasoz Cataner (Senior Creative Director, Zynga), and many more. Worth hitting play if you haven’t.

Sett is building agentic AI that helps mobile game teams find winning creatives faster and grow their games into global hits.

Listen to Unpredictable Hits

Top News Items

Written by Devin Becker, Consultant at Naavik

Character.AI wants a piece of the microdrama pie

• Character.AI, the company behind the popular AI chatbot service, launched “c.ai Series,” a slate of original short-form vertical dramas produced by an in-house studio with credits spanning Nickelodeon, DreamWorks, Netflix, and Blumhouse. The first titles include a romance anime, a Gen Z paranormal horror, and a virtual-world survival thriller, all produced using AI. The hook is that after watching an episode, 18+ users can open a chat with the characters and continue the story interactively. It is a clever retention play in a booming market. Short drama apps generated significant IAP revenue in 2025, with some expecting it to more than double in 2026 (see our recent digest and podcast coverage on microdramas). Character.AI is betting that layering conversational AI on top of vertical video creates a stickier loop than passive autoplay, one where the character, not the episode, becomes the durable asset.

• The question is whether this is primarily a content strategy or a platform strategy. Character.AI says it will eventually open production tools to creators, turning what is currently a studio-led slate into a broader ecosystem. Traditional microdrama economics are brutal, with as much as 90% of platform budgets reportedly going to marketing and user acquisition rather than content. If Character.AI can use its existing chat infrastructure and user base to bypass those acquisition costs, it could build a differentiated position. But first it needs to prove that chat-after-watch is a retention driver, not just a novelty. Fox, Peacock, and Disney are already circling the microdrama space with traditional IP. Character.AI’s potential advantage is that its characters can actually talk back.

Roblox Build enables creators to build one-shot games with AI

• Roblox announced “Build,” a mobile-first feature launching July 28th in New Zealand that lets any user turn a text prompt into a playable game directly within the Roblox app. Powered by a mix of open-source and proprietary Roblox AI models (including Cube, its 3D foundation model), Build handles gameplay mechanics, environments, characters, visual style, and sound. Users can iterate, playtest, and publish without ever opening Roblox Studio, though the two tools share a backend, so creators can start in Build and refine in Studio. This is a meaningful expansion of Roblox’s AI creation stack beyond the Studio-based agentic tools it shipped earlier this year. Published games will be available globally to age-verified users 16 and older and will be subject to Roblox’s existing retention-based discovery ranking, the same system that surfaces all other experiences on the platform.

• The slop question is unavoidable with one-shot development. Roblox is betting its existing discovery algorithm provides a natural quality filter: if nobody plays it, nobody finds it. That logic held when supply was constrained by the difficulty of building in Studio. Whether it holds when potentially millions of casual users can generate and publish games from their phones is a genuinely open question. The deeper strategic play is clear, though. Roblox wants every one of its 132M DAU to see themselves as a potential creator, not just a player. If even a small fraction converts, this helps discourage creators from considering using other creation platforms, especially those leveraging AI that are starting to emerge. Roblox will certainly be counting on social sharing when everyone can create something they can at least convince a friend to play.

Meta has released an app for making generative AI games

• Meta soft-launched Pocket, a standalone mobile app that lets users generate and share small interactive games and experiences using text prompts. The app appeared on iOS and Android on June 29th with no formal announcement, and it is not yet available in the US. Pocket is built on the bones of Gizmo, a vibe-coded gaming platform whose team Meta acquired earlier in 2026. Users can create mini-games from prompts and browse a scrollable feed of others’ creations. Gizmo had generated 635K installs with 98% positive sentiment before the acquisition, and Meta is looking to scale that concept with its distribution muscle.

• This is the latest in a string of experimental AI content apps from Meta (following AI-generated images via Meta AI, AI video via Vibes, and AI editing tools in Edits), but Pocket is the first one that sits squarely in interactive entertainment. The quiet launch suggests Meta is still testing product-market fit, and there is no indication yet of integration with Instagram, Facebook, or WhatsApp. But the potential distribution advantage is enormous. If Pocket’s gizmos eventually surface inside Meta’s social feeds, it could reach billions of users overnight. For the gaming industry, Meta entering the vibe-coded games space is noteworthy. This means the largest social media company views AI-generated interactive content as a product category worth investing in.

Yield Guild Games Launches vibecode.game and a Game Jam

• Yield Guild Games (YGG), once a standard-bearer of blockchain gaming, launched vibecode.game, a discovery and publishing platform for vibe-coded browser games. To kick things off, YGG is partnering with Minds by Animoca Brands for a four-week game jam running through August 10, which includes a “Game Designer Mind” AI agent that helps creators with design decisions, mechanics, and balancing. The pivot is notable: YGG and Animoca were among the loudest voices in the play-to-earn era, and this move signals a pragmatic migration from blockchain infrastructure to AI-powered creation tools. The platform features curated rankings, reviews, and community-driven discovery, positioning itself as a permissionless launchpad where anyone can publish and find an audience.

• The timing is interesting, as the bottleneck has shifted from production to distribution. Making a game is easier than ever, but getting anyone to play it is harder. vibecode.game is a direct attempt to build a discovery layer for the emerging wave of AI-built games. Whether YGG’s community, which was built around token-incentivized gaming guilds, translates effectively into a vibe-coded game audience is uncertain. YGG is trying to leverage its Web3 start to improve discoverability for AI-built games, but with such a small audience in comparison to a mature platform in the space like Roblox, this feels more like a last-ditch pivot as Web3 gaming has mostly died out.

Other News Items

• Meta says its new AI model is ready to compete on coding

• MIRA Multiplayer Interactive world models with Representation Autoencoders

• Kaon AI raises $60M for GenAI entertainment

• Grok’s programming tool was uploading its users’ entire codebase to cloud storage

• Moonshot’s Kimi K3 pushes Chinese AI into Fable-level territory

• FL Studio 2026 turns its AI chatbot into your assistant engineer

• Apple’s failed self-driving car program left a legacy of powerful AI chips

Content Worth Consuming

• Reverse Engineering a Commercial AI Cheat: How AI Agents Changed the Way We Work (Armada): “While my findings on the cheat are obviously an important part of this engagement, this methodology turned out to be just as interesting. Agentic AI tooling applied to reverse engineering is not a marginal improvement. The economies of large-codebase analysis have shifted. A piece of commercial software that might have required weeks of manual static analysis to map structurally can be characterized in days when you’re asking questions rather than reading files.”

• The State of AI in UA Creative (Sett): “Across the pipeline, AI now does the work that used to eat a creative team’s week with many creative tools that make it easier or faster to produce creatives - be it assets, optimization, editing, and more. It is also able to spin up variations of the same concept at a lower and lower cost - similar to templates. But creative tools have the same ceiling. They run on a concept you already have.”

• What do Steam fans really think about AI in games? (GameDiscoverCo): “Conditional acceptance is the order of the day: ~51% of the 2,100 analyzable answers accepted AI for some uses but not others, for example: ‘I’m fine with limited AI use for support tasks, not for replacing human creativity.’ On the ‘acceptable’ side, the clear winner is coding/programming: code helper tools like Claude Code had ~239 mentions as acceptable. This was way more than other categories like prototyping & and placeholder assets (119 mentions), and/or tedious or repetitive tasks (101 mentions).”

• Building an AI game testing agent with Amazon Bedrock (AWS): “Automated game testing isn’t new, but previous approaches have always struggled to keep up. Can an AI agent on Amazon Bedrock deliver autonomous QA that actually works?”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s how we can help, spanning all platforms, genres, and regions:

• Strategy Consulting: Market research, corporate strategy, game and economy design, gamification, live-ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above to learn more. If you'd like to discuss how Naavik can support your team, click the box below or send us a note at consulting@naavik.co.

Over the past weeks, Naavik covered the fast-growing microdramas category. If you’re paying attention to this space, we recommend checking out our podcast discussing AI-native production with Angela Yu, co-founder of Storeel, as well as our in-depth article overview.

As always, Naavik’s strategy consulting, M&A advisory, and fractional talent teams are here to help your business. Reach out to learn more.

Onto this week’s article, let’s dive into the AI chatter around Steam Next Fest.

Sponsor: Create Next Generation Game Experiences with AWS

AWS for Games gives you everything you need to build, run, and grow innovative gaming experiences. Leaders across the gaming industry rely on AWS’s more than 16 years of games expertise to launch and run some of the world’s biggest games, including Fortnite, Roblox, and League of Legends.

AWS and over 300 AWS Games Partners offer the broadest and deepest set of purpose-built services and pre-built solutions for cloud game development, content creation, game backend and infrastructure, player acquisition, retention, and monetization. These tools empower game developers to build healthy global player communities, gain player insights, personalize, and monetize player experiences, to build lifelong fans. 750 million gamers each month, around the world, play games built on AWS.

Learn how AWS can help you build, run, and grow innovative gaming experiences.

Learn More

Steam Next Fest’s AI Problems

Written by Francois Courset, Lead Consultant at Naavik

The June 2026 edition of Steam Next Fest closed out two weeks ago and was by far the biggest yet: over 4,300 demos for upcoming PC titles, with games like Mistfall Hunter, Echoes of Aincrad, and XenoFeels standing out as the event’s clearest breakout hits. What defined this edition, however, was the volume and intensity of criticism surrounding AI’s role in game development.

Because Steam Next Fest is an open window on what the PC landscape will look like for months to come, it’s also one of the clearest available signals for tracking AI adoption and consumer reception at scale. We wanted to zoom out on the event and assess its relevance when barriers to development are diminishing, unpack how studios should be reading the current backlash, and finally, use our results to map where AI-enabled content actually stands today.

Discoverability Challenge in an Increasingly Crowded Event

A quick primer on Steam Next Fest: It is Valve’s flagship demo showcase, giving players free and direct access to hundreds of unreleased titles. Launched in 2019 as the Steam Game Festival, the event now runs three times a year — in February, June, and October — with the June 2026 edition taking place June 15–22. Participation is limited to games with a Steamworks account in good standing, a functioning demo ready at launch, and a release date scheduled after the Fest wraps, with one-and-done eligibility per title. Over the years, it became a key beat for PC developers to collect early feedback, accumulate wishlists, and build pre-launch momentum.

Steam Next Fest just hosted its biggest edition yet: 4,300+ entries, a 24% jump from February and 66% growth YoY. The surge tracks the event’s well-established value for developer visibility, but it’s really a downstream effect of Steam’s expanding catalog, with new releases up 19% YoY in H1 2026 per Video Game Insights. AI has been a widely documented accelerator here. AI-tagged releases (a self-declared field) have climbed to as much as 40% of new titles in the most recent months, a direct consequence of AI tools lowering the barrier to entry for developers.

Attention at Steam Next Fest is increasingly becoming a zero-sum game. While the participant pool spikes, aggregate followers have increased just 7% from the last edition and 22% YoY, nowhere near enough to keep pace with the growth in new games. The result is smaller gains for individual games, as reported in Gamediscover.co’s extensive breakdown: top 10% performers added +121 followers (down ~25% from +163 in June 2025), and top 1% performers added +1,330 (down ~25% from +1,759). This means fewer breakout hits driven by the event, and instead more attention focused on games that had already built momentum through separate marketing initiatives.

Steam Next Fest is a live case study in the discoverability dynamics we unpacked in our Distribution in the Age of AI piece: too much content, not enough platform-level curation to sort it. Valve's slow pace of adaptation shifts the burden onto developers. Explosive submission growth is eroding Next Fest’s value as a standalone play while boosting the appeal of tightly curated alternatives, hence the rise of showcases like Evil Empire’s III initiative. Algorithmic placement alone no longer moves the needle. Winning now requires building an audience before the demo drops and focusing budget behind channels outside Steam’s oversaturated surface — press, influencers, community. Smaller teams without the budget to run that kind of funnel will find breakout outcomes harder to reach.

Navigating Gamers’ Concerns Over AI Adoption

In parallel, the event was marked by consumer resentment around AI disclosures, which sparked review-bomb campaigns further reinforced by media narratives. Industry leaders are now staking out public positions against Steam’s AI labeling push, while a growing (if still niche) contingent of players has turned to browser extensions built specifically to filter AI-tagged games.

So what exactly happened? Of the 4,382 demos featured during the event, 550 — or 12.5% — disclosed some use of AI in their production, a self-reported figure that likely understates true usage. That’s a far cry from the 20% number that made headlines, which actually referred to submissions. It suggests that Steam’s eligibility barriers are doing a good job of filtering the pure AI slop content.

The backlash, however, clustered around developers concealing AI use (such as 1666: Amsterdam). Several of the event’s top-performing demos disclosed their AI usage upfront and performed just fine. Six out of the top 25 games by wishlist growth had the AI disclosure.

This isn’t the first time an innovation has been branded anti-player. Microtransactions took close to a decade to go from pariah to standard practice on PC and console. We are clearly standing in a transition phase, and it’s on developers to thread the needle between production pressures and audience sentiment. Specific geos, such as Asian markets, have already achieved smoother consumer adoption.

A sticking point, however, is Valve’s tagging system, which treats wildly different AI use cases as a single undifferentiated category, artificially splitting the market into “AI” and “No AI” camps. A study by Ross Burton provides empirical evidence that the Steam AI tag correlates with weaker performance. This creates a fundamentally flawed incentive structure rewarding silence over transparency. It is clear Valve’s current system either needs to be radically revised or removed entirely.

Consumer resentment around AI is unavoidable simply because the loudest critics are also the most engaged gamers, the ones who drive initial word of mouth. The 1666: Amsterdam case makes clear that hiding is not a proper option, and developers might still underestimate the risks. To that end, the first step is proper governance: auditing use cases, vetting vendors, and setting disclosure rules before a title reaches Steam. The second step is communication: leaning into Steam’s disclosure field honestly while getting ahead of backlash, much like SIE has done by publicly contextualizing its own AI usage. As Steam Next Fest’s results show, some demos disclosing AI still met performance expectations, which suggests AI is frequently used as an argument for a game’s other shortcomings, while quality remains, as ever, the deciding factor for consumers.

Where Are AI-enabled Games?

The most notable absence in this edition of the Next Fest was AI-enabled games, meaning games that use generative AI as a core gameplay component — through AI NPCs, live content generation, or prompt-based interactions. None of the top 50 Steam Next Fest games by wishlist increase featured an AI-enabled gameplay component, and these types of experiences are barely part of the conversation.

It’s worth caveating this statement: Next Fest audiences aren’t especially AI-friendly right now, and running inference for free against a week of unpredictable, high-volume traffic is a rough economic bet. The issue is also largely PC-specific, as we’ve previously noted how the mobile ecosystem was quicker to iterate around AI usage. Still, a small pocket of demos experimented with AI-native mechanics, even if none broke out of niche territory. AI Pixel Battle, for instance, has players draw a monster tough enough to outmatch the opponent’s, then hands judgment to an AI that narrates the battle in real time.

We are still quite early in the generative AI cycle. It’s been less than 4 years since ChatGPT launched, and it takes time for specific tooling to be built, for companies to adopt it, and then to ultimately ship years later. Teams still often have to create their own tech, as no out-of-the-box AI tools exist flexibly at scale yet for gameplay-related use cases. Next Fest gives a glimpse into the near future, but is not fully representative of anything more distant. It isn’t leading to unique AI breakouts yet (and isn’t designed to), and it will just take more time to get to that point. We’ll just have to see how long it takes.

Looking at recent announcements from Krafton and Mihoyo, we could be witnessing an evolution where AI-gameplay innovation is pushed from the top, instead of emerging from startups that typically have more room to innovate but have to rely on available off-the-shelf solutions. By nature, AI technology can be complex to integrate and maintain in player-facing experiences. It requires specialized expertise, extra infrastructure, continuous tuning, and strong pipelines for data, testing, and model maintenance. At the same time, the economics of inference costs haven’t been fully solved.

What we’re witnessing right now is AAA publishers with enough runway to innovate going to market through measured innovation as optional aspects in already live games, while more experimental takes are either still largely in the proof-of-concept stage or live within isolated ecosystems like the web or Discord. Looking ahead, it’s genuinely hard to say where the next AI-enabled hit will emerge from, and we’ll be watching this space closely as it develops. In the meantime, even the first steps toward production-grade AI adoption remain genuinely difficult within PC. As the broader ecosystem transitions toward wider adoption, a fair amount of consumer education will be needed to bring players along, especially in the West.

Top News Items

Written by Devin Becker, Consultant at Naavik

Godot Bans "Autonomous AI Agent Use or Vibe Coded" Contributions

• The Godot Foundation announced policy updates banning nearly all AI-generated code contributions to the open-source game engine. The updated rules prohibit autonomous AI agent use or "vibe coding" — using AI to generate substantial portions of code — as well as AI-generated text in human-to-human communication, such as pull request discussions. Minor AI assistance for tasks like code completion, regex, or find-and-replace remains acceptable, but contributors who use AI to help author code must disclose it. The foundation’s reasoning is blunt: AI-generated contributions have been increasing, and they are “demoralizing” for the volunteer maintainers who review them. Godot wants all contributions made by humans who can “take responsibility for their code,” and heavy AI users often won’t do that.

• This is an important line drawn by a major engine project, and it matters because Godot’s open-source codebase depends entirely on volunteer goodwill. If maintainers burn out fielding low-quality AI submissions, the engine suffers. The policy highlights a growing philosophical split in game engines as commercial engines like Unity and Unreal are racing to embed AI deeper into their workflows. Godot’s bet is that human accountability produces better infrastructure code, while the commercial engines bet that AI acceleration is an unavoidable competitive advantage. It’s important to note that Godot isn’t preventing developers from using AI coding tools on their own projects, only preventing contributors from submitting that code back to the engine. This stance may also appeal to developers who have previously considered Godot when alienated by Unity’s past policy changes, but it could also lead to an increased feature gap over time.

General Intuition Raises $320M for AI Frontier Models Based on Gameplay

• General Intuition, a sister company to gaming clip platform Medal, raised $320M at a $2.3B valuation to build AI frontier models trained on gameplay data. The pitch is that games generate uniquely valuable spatial and temporal reasoning data that can train AI models to perceive, predict, and act in any environment, including controlling real-world robots. The company builds models that predict actions, like the decisions players make inside video games. Medal’s 17M+ monthly active players provide the raw data, and the company has already transferred its learnings into real-world robotics faster than expected.

• The strategic significance lies in gaming data being valued as foundational training material for general-purpose AI. If General Intuition’s approach works, game data becomes an important upstream input for the broader AI economy. Spatial data from other companies, like Niantic, has already proven interesting to AI companies, and any potential revenue source for games is worth considering in today’s constrained environment. For more on General Intuition’s approach, listen to our interview with Pim de Witte here.

Kinoa Raises $10M to Unlock Mobile App Revenue Using AI

• Kinoa raised $10M to support its suite of AI predictive models for mobile app operations. Founded by veterans of Playtika, Amazon, and Skai, the platform uses predictive AI agents to anticipate which users will churn, become high spenders, and which offers to surface at the right moment. It then acts on those predictions automatically without requiring code releases. The company claims it consistently delivers over 25% revenue lift across its customer base, which includes Playstudios, Playsimple, and MTG. The pitch is particularly timely given that mobile user acquisition costs have increased while targeting efficiency has declined, squeezing the economics that mobile publishers have relied on for years.

• While most AI conversation in the games industry centers on content creation (generating art, code, or narrative), Kinoa targets the operations layer — live ops, monetization, and retention. This is an area where AI can deliver fast and measurable ROI because the feedback loops are tight and the outcomes are directly tied to revenue. If tools like Kinoa prove out at scale, they can allow small mobile teams to run personalization programs that previously required a large dedicated live ops staff. Such tools can also boost the efforts of larger teams, although they may potentially compete with legacy in-house tools.

Savvy Games Group Teams with Genvid to Bring AI Tools and Training to Saudi Arabia

• Saudi Arabia’s Savvy Games Group signed a memorandum of understanding with Genvid Holdings and Massive Studios to bring AI-powered development tools, training programs, and mentorship to Saudi-based game studios, indie developers, and universities. The deal focuses on Savvy’s incubator efforts and job creation within the kingdom, not on Savvy’s larger portfolio companies like Scopely or ESL FACEIT. Genvid will provide its enterprise AI workflow platform, which includes provenance tracking that can verify the origin of generated assets for copyright compliance, and it will run master classes teaching local developers how to build cutscenes, assets, and trailers using AI. Massive Studios will contribute its AI-native production capabilities for visual storytelling.

• Genvid CEO Jacob Navok argues that AI game tools may take off outside the United States first. He pointed to the absence of anti-AI labor pushback in regions like Saudi Arabia, India, and Southeast Asia, and noted that these markets have access to inexpensive labor that can be combined with inexpensive AI tooling to produce content at scale. Saudi Arabia’s Vision 2030 initiative pushes building a domestic games workforce quickly, and AI tools can compress the learning curve for developers who lack the decades of institutional knowledge that established Western studios have accumulated. If the approach works, it could become a template for other emerging game ecosystems and a competitive pressure on established studios that are slower to adopt these same tools.

EA: AI Is Delivering "Faster Prototyping" and a "Real Rise in Creativity" in Its Studios

• During Summer Game Fest, EA’s president of enterprise development Laura Miele said AI tools are producing tangible benefits inside EA’s studios, including faster prototyping, reduced tedium in developer workflows, and what she described as a genuine increase in creative output. She did not cite specific examples but positioned AI as a friction-removal layer for pipelines and tools rather than a replacement for creative labor. This follows EA’s 2025 directive urging nearly 15,000 employees to lean on AI for tasks ranging from code generation and concept art to management workflows, and QA automation. Back in 2024, CEO Andrew Wilson estimated that more than 50% of EA’s development processes could be positively impacted by generative AI.

• Miele’s comments are notable less for what they say and more for what they don’t. No shipped titles were referenced, no metrics were shared, and the specific types of AI being used were left undefined. There is a wide gap between AI-powered project management tools and generative AI content creation, and Miele’s remarks did not distinguish between the two. That vagueness is a pattern across major publishers where executives signal enthusiasm while the actual integration remains opaque. The real test will be whether EA’s upcoming slate (Battlefield, sports titles, and any new IP) visibly reflects these gains. Until concrete results ship, these statements read more as positioning for investors and the pending Saudi acquisition than as evidence of a creative transformation. However, if the results do materialize, it will likely further accelerate the adoption of AI tools in larger publishers.

Other News Items

• In World of Warcraft, a server without humans has appeared - instead, 1,800 DeepSeek-based bots are playing there

• Tripo AI raises $150M for GenAI tools for gaming — a month after its previous $200M raise

• Vampire Survivors dev ‘reviewing’ Fortnite collab after Epic’s AI announcement

• Netflix is using an AI-generated Gene Wilder voice in its Willy Wonka reality show

• Figma now has AI motion graphics and shader tools

• Photoshop and Premiere now have AI assistants

• NexTide Media launches AI tool to glean audience insights from livestreamed video

Content Worth Consuming

• Epic’s Sweeney claims Steam AI labels are ‘really irresponsible of Valve’ (GamesIndustry.biz): “I think it’s really irresponsible of Valve. They shouldn’t do it, because it makes it much, much, much harder for a game developer to have a chance of success. You have to choose from either not using tools that can make you way more productive, and probably failing due to competition that does.”

• How Tencent sees AI and UGC lowering game dev costs and leading to new kinds of games (Gamesbeat): “Look, at the end of the day, I know in the West there’s a lot of resistance in the gaming industry in terms of thinking about what AI means for the industry as a whole. In terms of labor and everything like that. Working for Tencent, I think you know Tencent’s a very practical company as well. When I went back to Shenzhen a few months ago, management talked a lot about AI first across all of the business.”

• AI Native Games: A Survey and Roadmap (Zhongguancun Institute of Artificial Intelligence): “Generative AI now enables games to produce dialogue, quests, characters, images, and worlds at runtime. Yet generation alone does not make a game AI-native, nor does it guarantee playability. This paper defines AI-native games by whether runtime generative AI is constitutive of the core loop: if the AI component were removed or trivially replaced, the central form of play would collapse or become fundamentally different. This counterfactual criterion separates AI-native games from AI-augmented games, boundary artifacts, chatbots, tavern-style roleplay, procedural content generation, and AI-assisted production. Using this definition, we screen candidate artifacts and analyze 53 publicly available AI-native games and prototypes.”

• I Gave an AI a Civilization to Run. It Built a Nuke (Liam Wilkinson): “I gave an AI a civilisation to run. By the midgame it was winning: a trade network that dominated the map, alliances on every border, a diplomatic victory within reach. It had outbuilt, outearned, and outmanoeuvred every rival on the board. What it hadn’t noticed was France. Quietly, across a hundred turns, French culture had been seeping into every city on the map. By the time the agent recognised the threat, the tourism was so deeply embedded there was no peaceful way to stop it. Every counter it reached for was broken. Every tool it had built to respond failed. It had one option left. It built two nuclear devices and levelled Toulouse.”

• MIRA: Multiplayer Interactive World Models with Representation Autoencoders (General Intuition & Kyutai): “Simulating video games is therefore a stepping stone to physical AI: it's a setting where data is abundant and it's cleaner than real-world videos since it's generated by a game engine. But learning the mechanics of an AAA video game purely from observing video is still no easy feat, so the hope is that what we learn by studying video game world models will be useful in real-world settings later.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, live ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.

This week, we dive into the emerging category of world models and explore the impact they could have on the gaming ecosystem. For additional context, we highly recommend our recent interview with General Intuition’s CEO, where he discusses the potential of world models.

Additionally, our partner Sensor Tower just released its State of AI 2026 Report. Go check it out!

Sponsor: How Agentic AI Drives Efficiency in Game Development

Industry leaders from Xbox, Sega, PlayStation Studios, and EPAM recently gathered to examine what happens when AI stops following scripts and starts shaping experiences. The conversation covered autonomous NPCs, adaptive world-building, AI safety, and the critical role that human creativity still plays when technology handles repetitive work.

Key themes explored:

• From copilot to agent. AI has evolved from code-assist tools to end-to-end agentic systems capable of autonomously delivering entire features.

• Creativity amplified, not replaced. AI accelerates prototyping, but taste and craft remain irreplaceable — tools alone never reach a creator’s standard.

• Living, infinite worlds. Games can now adapt in real time to a player’s personality, mood, and history — moving from static narratives toward endlessly evolving content.

• Safety first. Player-facing generative AI must be load-tested like live services; poor personalization is “toxic to the experience.”

• New genres emerging. AI-native games and text-based formats are making a comeback, giving players deeper agency.

Read the full report here: Agentic AI in Gaming: Evolving Game Development and Player Experiences | EPAM

Read the Report

World Models' Transformative Role in Gaming

Written by Francois Courset, Lead Consultant at Naavik

World models have emerged as one of the most engaging technology narratives both within and beyond gaming. From Google Project Genie’s preview triggering a shockwave across gaming stocks in February 2026 to a recent wave of high-profile announcements — including Roblox’s acquisition of Morpheus AI, Tencent’s release of HY-World 2.0, AMI Labs’ $1B raise, and Tripo AI’s $200M funding round — it is easy to feel overwhelmed.

Now is a good time to take a step back and synthesize all the action and noise. Why are we witnessing accelerating investments into world model companies? What disruptions may the gaming industry face? And what role will gaming companies play in shaping the burgeoning world model ecosystem, which touches far more than gaming itself?

What Are World Models?

World models are AI systems that learn from video, simulations, and other spatial data to build internal representations of scenes and objects, enabling them to predict how an environment evolves and how an agent’s actions change it. Given a current state and an action, they can predict the next state in a way that stays consistent with the rules of the world they learned from. It’s all about modeling action-induced cause and effect. They differ from video models (Sora, Veo, etc.) that produce plausible footage from a prompt without a structured action input.

For a more detailed introduction, we recommend looking at Naavik’s deep dive on world models, Not Boring’s in-depth essay, and MoE capital’s overview which we’ll refer to throughout this article.

Not all world models are created equal. Xun Huang, CEO of Morpheus AI (recently acquired by Roblox), argues that five properties separate world models from traditional video models: causality, interactivity, persistence, real-time responsiveness, and physical accuracy. Models such as Genie 3, DreamDojo, and DreamZero each make different tradeoffs across these dimensions.

World models are a recent breakthrough, because they finally merge two research traditions that each had half the solution. One tradition built AI that “dreams” — learning a model of the world and practicing actions inside its own imagination — but these couldn’t generalize or produce realistic visuals. The other learned from massive amounts of internet video to generate photorealistic, physics-obeying footage (like Sora) but wasn’t interactive; you couldn’t feed it an action mid-stream. Between 2024 and 2026, advances made video generation fast and action-responsive enough to combine both, producing simulators that are realistic and interactive.

This ultimately led to the first generation of world models that provided a glimpse at the technology’s potential, culminating with Google Genie’s preview in early 2026 (see Naavik’s dedicated breakdown).

The potential of world models spans far beyond gaming. While world models use gaming videos as training data and grounded their earliest demos in gaming worlds (like Decart’s Oasis with Minecraft or GameNGen with Doom), the technology’s potential spans multiple industries. Key application areas include robotics, autonomous vehicles, military, training and simulation, architecture, entertainment, and, of course, gaming. It’s key to note that the potential in other areas — like robotics — is likely far larger than gaming. To address this breadth of use cases, different groups of world models are emerging:

• Renderers: models that take in actions and output visually realistic pixel observations for humans, focusing on appearance rather than true 3D structure or accurate physics.

• Simulators: models that output underlying world state (geometry, physics, dynamics) so both humans and programs can reliably inspect, compute on, and interact with a structurally accurate environment.

• Planners: models that take observations and goals as input and output sequences of actions, deciding what an agent should do next and thus closing the perception–action loop.

World Models’ Takeoff Phase

The past six months have marked an acceleration in investment and announcements tied to the world model ecosystem, pushing the category into one of the most aggressively funded frontiers in technology. More than $10 billion has flowed into the category over the past 18 months, including funding rounds such as AMI Labs at $1.03B, World Labs at $1.23B, Runway north of $860M, Rhoda at $450M, and Decart at $153M, and likely more to come very soon.

Gaming companies are piling into the world model race, too. General Intuition raised $134M built atop Medal’s vast library of gameplay clips, 3D asset generation startup Tripo pulled in $200M to push its own world model ambitions, Roblox stood up a new Roblox Reality division by acquiring Morpheus AI, and Krafton stretched its AI bets all the way into robotics and military simulation.

Taken together, these point toward the opening act of a hype cycle reminiscent of the LLM surge that detonated in late 2022. Capabilities are improving rapidly, unlocking new use cases and accelerating adoption.

This is precisely where gaming professionals need to grasp the full scope and ambition of world models: those $100M+ seed rounds and billion-dollar valuations simply can't be justified by gaming applications alone. To justify these valuations, companies will likely need to reshape larger industries such as robotics or help create entirely new forms of interactive experiences.

From Frontier Labs to Structured Ecosystem

Before diving into gaming specifics, let’s figure out what’s actually happening behind all of the headlines.

First is recognizing that world models have moved past the pure-research phase, with the first commercially-ready, open-source models — Nvidia’s DreamDojo and Tencent’s HY-World 2.0 — now widely available. This pushes AI labs to differentiate less on whether the tech works and more on inference economics and on acquiring quality training that sets their models apart.

Next is specialization. From narrowing focus to one of three model categories (Renderers, Simulators, or Planners), to building specialized models trained on domain-specific data sets. The next generation of models will move beyond proof-of-concept demonstrations and become increasingly optimized for specific use cases such as generating scenes and supporting actions in 3D spaces.

Lastly, and perhaps most importantly, world model initiatives are expanding beyond foundational models and are shaping a comprehensive stack. Stable-worldmodel provides a unified, modular framework for the entire pipeline of world model development: data collection, training, and evaluation. Reactor World positions itself as an infrastructure layer for the “World Model era,” where every major world model is available on one API. As happened with the LLM ecosystem, new entrants are likely to emerge across every layer of the value chain, from consumer-facing applications to the infrastructure and hardware layers. What we’ve seen across the first half of the year is the emergence of the first contenders in this emerging “world models stack.”

What’s Next for Gaming Companies

Arms race for talent at the intersection of 3D and AI

If world models are the next frontier in AI, the people best equipped to build them may already be sitting in game studios. World-model researchers occupy one of the thinnest slices of an already supply-constrained AI talent market, and gaming professionals have spent decades solving exactly the kinds of problems these systems require: physics, 3D environments, and interaction dynamics within simulated worlds.

It is no coincidence to see acqui-hire announcements in and around gaming startups, from the founders of Morpheus AI, Dynamics Lab, Lucid AI jointly joining Roblox, to Google Deepmind acquiring three startups at once, including 3D asset-generation startup Common Sense Machine.

While these acquisitions offer healthy exit opportunities for startups still wrestling with product-market fit, they also nudge gametech founders away from the problems that genuinely matter to our industry, such as pipeline-ready asset generation. This creates a paradoxical innovation squeeze, as key technical talent gets redirected toward larger non-gaming opportunities with deeper pockets.

Recognizing the value of gaming’s data

A world model learns the dynamics of an environment. It takes in observations — usually video frames — builds a compressed internal representation of the world’s state, and predicts how that state will change in response to an action. The key ingredient is the causal link between action and consequence, which can’t easily be scraped from the web. Gaming, it turns out, is sitting on the richest supply of exactly that through gameplay videos or engine-level capture.

Companies able to systematically gather and label gameplay data are bound to play a key role in world models’ future development. That’s the thesis behind General Intuition, which leverages Medal’s library of roughly two billion clips a year from around ten million monthly users across tens of thousands of games. The same goes for Microsoft and Minecraft’s years of first-party gameplay telemetry, Amazon with Twitch, or Google with YouTube’s data. Roblox could also play a significant role in world model development, because it possesses the world’s largest multimodal 3D dataset of proprietary user interaction data, paired with millions of 3D object assets and even user-generated clips.

More importantly, it reframes data from a supporting asset into a core value driver for gaming companies. Origin Lab, which provides a marketplace for capturing, creating, and delivering premium, rights-cleared multimodal content for AI training, could be a key enabler for companies to better monetize their proprietary datasets.

The pipeline efficiency perspective

The most immediate application of world models in gaming will be at the front of the pipeline, where they enhance ideation and compress prototyping cycles. By drafting environments, scenes, and dynamic storyboards on demand, teams can iterate on atmosphere, pacing, and composition long before committing to full asset development.

The next unlock sits further downstream in level and world design, as models generate persistent, editable 3D worlds from text, images, videos, or panoramas and export them in standard game-engine formats — a use case World Labs is actively chasing with its Marble product.

The longer-horizon and more bullish case is Roblox’s approach, applying world models inside UGC ecosystems to deliver photorealistic output without manual high-fidelity asset creation. That would meaningfully lower the barrier to entry, expand Roblox’s reach, and let it compete more directly with traditional AAA experiences.

Consumer-facing innovation is still far on the horizon

We have already shared our doubts on the game engine-killer vision for world models when Project Genie was first shown, simply due to games being defined by stable, repeatable rules and deterministic outcomes. Rather, perhaps there is potential in a hybrid entertainment experience somewhere between games and videos. There are reasons to believe this prospect remains far on the horizon.

Nearly four years into the generative AI wave, we’re only now seeing the first iterations of AI that deliver meaningful, genuinely fun value to the player experience — and world model–based innovation is likely to follow a similarly long experimentation phase. That timeline is compounded by the need to craft entirely new interfaces for engaging with world-model experiences, a challenge the VR ecosystem knows well. Cost is another constraint: today’s world model inference runs an order of magnitude higher than comparable workloads, with Genie 3 estimated at roughly $100 per hour, and while foundational model iteration should bring that down, solving the cost equation will remain central to any consumer-facing application. Finally there’s the question of demand. A new kind of experience may be technically possible, but VR is a cautionary tale that content innovation doesn’t automatically translate into consumer pull, let alone market share against traditional gaming.

The billion-dollar rounds aren’t being underwritten by gaming, but gaming data still has the potential to help catalyze the transformation of other industries. Gaming is historically at the tip of the technological spear (3D graphics, online interactivity, etc.), and its uses often set the foundation for progress elsewhere in the world. World models are no different. Although the future is fuzzy in terms of how world models may transform entertainment experiences, progress elsewhere looks undeniable and exciting. Of course, we’re incredibly early in the journey of world models and are excited to track its progress, inside and outside of gaming, over time.

Top News Items

Written by Devin Becker, Consultant at Naavik

Krafton Launches Beta for AI Ally in New PUBG Arcade Mode

• Krafton opened a beta test for Ally Duo, an arcade mode in PUBG: Battlegrounds where players team up with Ella, an AI companion powered by Krafton’s PUBG Ally technology. Built on NVIDIA ACE and running an on-device small language model, Ella responds to voice commands, understands in-game situations, and adapts her behavior in real time, covering movement, item collection, and combat strategy. Unlike traditional bot teammates that follow fixed scripts, Ella processes voice input alongside the evolving match state to act more like a reactive human partner. Krafton first announced the PUBG Ally concept in October 2025, but implementation took significantly longer than expected.

• The extended development timeline is worth noting because it underscores how difficult it is to ship AI companions that feel genuinely useful in a fast-paced, competitive environment. Battle royale matchmaking has long struggled with unreliable teammates, so a dependable AI partner could address a genuine player pain point. If Ella performs well enough, this could become a template for how live-service shooters help retain solo players who churn out of squad-based modes. Running on-device rather than in the cloud introduces potential hardware limitations, but it also helps keep costs manageable — an important consideration for a free-to-play game operating at PUBG's scale. See our previous coverage of Krafton’s AI strategy here.

Mobile Game Neural Dawn Showcases Arm's Latest AI and Graphics Technology

• Arm and Sumo Digital unveiled Neural Dawn, a mobile game built to demonstrate Arm’s Neural Technology running on next-generation Arm-developed Mali GPUs. It is the first mobile game to use Unreal Engine MegaLights in real time, delivering ray-traced shadows, hundreds of dynamic lights per level, and cinematic-quality rendering within a mobile power budget. The key enabling tech is Arm’s neural accelerators built directly into the GPU, running Neural Super Sampling and Denoising (NSSD) and Neural Frame Rate Upscaling (NFRU) to offset the cost of advanced lighting. A team of just 17 people at Sumo Digital built the game in roughly 18 months using standard Unreal Engine workflows and Arm’s plug-in tools.

• What matters here is more the proof point and less the game itself. Mobile has always been constrained by the tension between visual ambition and battery life, and Arm is demonstrating that neural graphics can potentially change that equation. If these capabilities ship at scale in consumer devices later this year as planned, they could materially raise the visual baseline for mobile games by bringing rendering techniques that remain uncommon even on consoles to a much broader audience. It remains to be seen what the install base will look like for those devices once shipped, and the games will still need fallback graphics. Regardless, game developers will no doubt showcase the games using this rendering tech.

Crystal Dynamics Uses Generative AI for Early Prototyping, but "Finished Content Is Human-Created"

• Crystal Dynamics addressed growing backlash over the AI disclosure on the Steam page for Tomb Raider: Legacy of Atlantis. Experience director Jeff Adams explained that the studio uses generative AI during early level design to quickly visualize candidate objects in a scene before committing developer time to building them. If the AI-generated visualization works, the asset moves into the traditional production pipeline where artists then build it from scratch. Adams stated that all finished content in the final game is human-crafted.

• Crystal Dynamics is trying to thread a needle on transparency. The “AI for ideation, humans for shipping” framing is becoming the standard talking point, and it will likely become more common as Steam’s AI disclosure requirements force transparency. Usually this gets talked about when teams are caught “accidentally” shipping AI assets in the finished product. The real tension is whether players and communities will accept any level of AI involvement regardless of how the final product is made. With unnecessary backlash from gamers surrounding non-player-facing uses of AI, this is another example of why Steam’s forced disclosure may be doing more harm than good without much more narrow usage.

Welevel Unveils AI-Based City-Building Survival Game SolidRiver and $8.5M in Funding

• Munich-based studio Welevel closed an $8.5 million funding round and revealed SolidRiver, a city-building survival game that uses AI to power smarter NPCs, dynamic gameplay, and developer productivity. Founder Christian Heimerl, a self-described hardcore MMO veteran, described the vision as creating a living world with AI serving as a real-time dungeon master. The AI handles tasks players find tedious (chopping 1,000 trees, for example) while also driving NPC behavior and world reactivity. Notably, Welevel is also building technology that could be licensed to other game studios.

• The dual-purpose approach — building a game and a platform simultaneously — is becoming a pattern among AI-native studios due to the lack of ready-made solutions. Whether SolidRiver succeeds or not as a game, the underlying tech becomes a possible middleware play. This strategy reflects this particular phase of funding AI game development, similar to what we saw in web3 gaming a few years ago. Many investors are drawn to startups where the IP extends beyond a single title or acts as a launchpad for licensing tech to others. The $8.5M raise is modest by AAA standards but meaningful for an indie studio betting that AI-driven world simulation is a defensible competitive advantage.

NVIDIA RTX Updates Bring DLSS 4.5 to UE5 and Multilingual AI Characters

• NVIDIA shipped a batch of RTX ecosystem updates aimed squarely at game developers. DLSS 4.5 is now available as an Unreal Engine plugin, delivering Dynamic Multi Frame Generation, a new 6x mode, and a second-generation transformer model for Super Resolution. Alongside the rendering upgrades, NVIDIA ACE expanded its multilingual AI character capabilities with three new models: Qwen 3.5 4B for low-latency dialogue across 201 languages, Riva Parakeet for speech recognition in 25 languages, and Chatterbox Multilingual for expressive voice synthesis in 24. All run locally on RTX hardware via the NVIGI SDK.

• The practical significance is the convergence of rendering and AI character tech into a single, integrated toolchain. NVIDIA is making it increasingly frictionless for UE5 developers to ship both better-looking games and conversational NPCs without stitching together third-party solutions. The multilingual angle is especially notable as localization of AI-driven dialogue has been a major open question, and running it on-device sidesteps the latency and cost issues of cloud-based alternatives. Studios that adopt NVIDIA's stack early could gain a head start in shipping AI characters to global audiences, assuming it provides meaningful advantages over third-party tools that may already be more mature. The easier this process becomes, the more likely we’ll see games experimenting with the technology despite some current gamer backlash.

Other News Items

• Sony AI Releases Woosh Foundation Model for Sound Effect Generation

• Crazy Taxi: World Tour Lead Says Generative AI Will Continue To Be A ‘Hot Topic,’ But They’re Using It Anyway

• Multiple clones of Minecraft and Pokemon made within hours of Claude Fable release before it was taken down

• Claude Code Game Studios turns a single Claude Code session into a “full game development studio”

• SpriteCook creates animated characters, sprite sheets, and tileable textures and adds them into your engine through a coding agent

• Autodesk’s Neural CAD brings AI reasoning to design and engineering

• ESA warns ‘No Fakes Act’ fails to distinguish between deepfakes and digital replicas used in video games

Content Worth Consuming

• Sensor Tower: State of AI 2026 Report (Sensor Tower): “The trend suggests that AI is becoming an increasingly effective discovery and user acquisition driver across app categories. As consumers actively seek AI-powered experiences, developers are gaining new opportunities to attract users by highlighting AI capabilities in app listings and marketing.”

• Take-Two’s former head of AI shares his concerns on the current hype cycle (GamesIndustry.biz): “’It’s so contextual,’ he says. ‘If you’re a tiny startup and you are going out of business in six months, why wouldn’t you use every advantage available? If you want to genAI all the things, it’s important that you understand the ethical implications of that. It’s a really hard one to wrestle with. If you have to pay minimum wage to bring in an artist, it’s already been an accepted practice to bypass minimum wage in your region and outsource it to somewhere where there’s a cheaper cost of living and pay below minimum wage for the same output. In some ways, how is this different?’”

• Arkadium launches GameLab to make AI ready for the real world (GamesBeat): ‘“I am having more fun now than I have in the last decade, and where most game developers are saying, ‘How do I use AI to make game production faster or cheaper, with my art or my coding?’ We’re flipping that whole thing on its head and saying, ‘How do we use games to make the models better, right?’” Rosenblatt said.’

• The devs of Chinese hit The Scroll of Taiwu on what Western devs get wrong in China - and why they rejected generative AI (GamesIndustry.biz): “Qiezi clarifies that no AI was used in the development of The Scroll of Taiwu. ‘But I think AI can be used as a tool and people are allowed to try it,’ he says. ‘Not in game creation, but more in management work or as a tool to reduce the burden of communications. For a game, it needs to be unique and to be something that people didn’t think of before or couldn’t imagine before, so those kinds of artworks could not be produced by AI.’”

• The future of Hollywood isn’t feeding prompts into vanilla gen AI models (The Verge): “Watching all of these films, I got the distinct sense that there is no future where studios are cranking out commercially viable projects by feeding prompts to gen AI models. That kind of content probably isn’t going to go away, but it’s not the kind of stuff Hollywood’s heavyweights would want to put their names on. What seems much more likely is bigger AI firms like Google partnering with studios to build bespoke models that are tailored to very specific workflows. And those workflows really only function well when they’re guided by human artists with very clear creative visions.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, live ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.

This week, we dive into one of the most mature use cases for AI: Ad Creative production, from Sett AI’s $30M raise to some of the current implementation strategies, as well as the long-term transformative potential for mobile games’ operations.

As always, we are interested in hearing from you, including your perspectives and experiences deploying these tools — let us know! Naavik’s team will also be attending VivaTech in Paris between June 17th and June 20th. Reach out directly if you’d like to meet.

Sponsor: Create Next Generation Game Experiences with AWS

AWS for Games gives you everything you need to build, run, and grow innovative gaming experiences. Leaders across the gaming industry rely on AWS’s more than 16 years of games expertise to launch and run some of the world’s biggest games, including Fortnite, Roblox, and League of Legends.

AWS and over 300 AWS Games Partners offer the broadest and deepest set of purpose-built services and pre-built solutions for cloud game development, content creation, game backend and infrastructure, player acquisition, retention, and monetization. These tools empower game developers to build healthy global player communities, gain player insights, personalize, and monetize player experiences, to build lifelong fans. 750 million gamers each month, around the world, play games built on AWS.

Learn how AWS can help you build, run, and grow innovative gaming experiences.

Learn More

The State of AI for Mobile Ad Creatives

Written by Francois Courset, Lead Consultant at Naavik

With fully AI-generated creatives flooding the market across mobile games of all sizes, it has become pretty clear that ad creative production has been one of the most promising verticals for AI adoption. In a 2025 survey from Appsflyer covering 700 mobile marketers, over half of the respondents reported that creative production and optimization were the two areas where studios benefited most from AI. As a multitude of specialized tools have emerged, Sett’s landmark $30 million Series B in March 2026 has further cemented interest in this area of application.

Why is there such a need for AI applied to ad creatives? How mature is the technology and associated toolset? Essentially, how will ad creative automation affect the broader mobile gaming landscape? Let’s dive in!

Ad Creatives Production Is Ripe for Disruption

For most of the last decade, mobile game user acquisition relied on finding the right user, but this has gotten considerably harder with privacy frameworks (ATT) gaining ground. Today, the balance has shifted to making the right ad — and making enough of them. Publishers have leaned more into "spray and pray" by testing and deploying a larger volume of ad creatives to maximize targeting surface area, in hopes of finding and acquiring the "right users." Production volume has exploded as a direct consequence. Currently, top gaming spenders are producing 2,400–2,600 creative variations per quarter, up 25–30% YoY.

This is further compounded by:

• Platform automation: The algorithmic side of mobile UA, which covers bid optimization, targeting, and budget pacing, has largely been absorbed by ad networks, leaving little room for publishers to manually optimize UA. Therefore, publishers are focused on the one variable they can still largely control — creative strategy.

• Creative trends acceleration: Creative isn’t a static medium. Winning formats cycle quickly (UGC ads, fake ads, mini-game hooks, etc.), forcing publishers to continuously test new formats to stand out from competitors. With competition increasing, more creatives hit the market, making trends evolve faster, while ad fatigue further compounds the need to refresh materials constantly. This creates a self-reinforcing loop, explaining the continuous spike in creative production.

• Localization: Rising requirements for localization to address secondary markets further amplify the need to expand production output. A single concept must be multiplied across formats (video, playable, UGC, static), channels (Meta, TikTok, AppLovin, Google, each rewarding different executions), and markets. A handful of “ideas” becomes hundreds of “assets” before a campaign even launches.

Creative production at scale has become a core competitive advantage, allowing teams to secure more attractive CPIs. In this context, AI-assisted production, faster iteration, and automation have become essential tools that enable smaller advertisers to compete more effectively and bigger ones to further expand output more cost-effectively. In digital video advertising, whose learnings can largely carry over to a mobile-specific context, 51% of advertisers already use AI-assisted workflows for digital video ad creation, with an additional 34% planning to use it. According to IAB, 39% of all digital video ads will be created using Generative AI this year.

AI creatives (where the video content is entirely AI-generated) have simultaneously emerged as a trend in itself, further accelerating tool adoption. 56 of the top 100 grossing mobile games used AI-generated ad creatives in 2025, with various implementations.

How AI Tools Support Ad Creatives Production

AI implementation in ad creative pipelines covers a variety of use cases, resulting in the entire production stack being AI-enabled. Key applications include:

• Ideation and Storyboarding: In line with any creative process, AI tools can help brainstorm new concepts, draft ideas, and generate early concepts and storyboards that can then be iterated upon.

• Localization: AI handles the translation of ad copy, automated voiceover generation across languages, and the adaptation of on-screen text and cultural references, allowing a single master creative to be duplicated for multiple markets.

• Resizing: AI automatically reformats a core asset into the many aspect ratios and dimensions that platforms demand, intelligently recomposing layouts so key elements (logo, call-to-action, character) stay in frame.

• Iteration: To feed the need for extensive A/B testing, generative tools spin up dozens or hundreds of variations of a concept by swapping sequence, hooks, color palettes, influencers used in UGC ads, music, copy, and CTAs.

• Video Asset Generation: Video creation now spans the full spectrum from AI-generated UGC-style ads (synthetic avatars, AI voiceovers, and script generation that mimic authentic creator content) to fully AI-produced cinematic spots built from text or image prompts.

• Playables Production: Playables can increasingly be generated and iterated on with AI assistance, which can generate gameplay hooks or replicate competitors’ mini-games faster than hand-coding each one.

• Insights Layer: Performance data from live campaigns (click-through rates, install rates, retention) is fed back into the system so the agents learn which creative attributes drive results, surface actionable insights and trend patterns, and power an ideation engine that proposes new concepts grounded in what's actually winning. This creates a continuous loop where every campaign more intelligently makes the next batch of creatives.

These use cases revolve around a core intelligence → generation → deployment → feedback loop that allows teams to generate multiple creative assets, as well as test, monitor, and identify winning concepts. A comprehensive range of solutions has emerged, tackling one or more of these use cases.

Key differentiation angles lie in three main vectors. First is the depth of stack coverage (e.g., Sett, Reforged Labs) versus more specialized creation tools (Poolday, PlayableMaker, MakeUGC) as showcased on the map below. Second is the user journey from a simple prompt-based interface (Poolday) to more hands-on workflows. Finally, the technological moat is typically concentrated in the agentic orchestration layer (Sett) as asset generation models tend to become commoditized.

What makes ad creatives uniquely suited to this new AI stack is the importance of iteration. Creative testing has always been a numbers game, and rather than marginal efficiency gains, AI workflows can now exponentially grow the output by an order of magnitude, which perfectly aligns with the A/B testing requirements in mobile UA. Reforged Labs claims to increase the speed of ad conception and storyboarding by 5x; Layer similarly claims the ability to produce 50-100 creative concepts per day per strategist, compared to 5-10 with traditional methods; Sett improved output quantity and quality for Supersonic, driving clear Share of Voice impacts; Creatify reports up to 97% time saved in creative production, 90% lower production cost, and up to 50x more videos (1,000+ creatives/month) vs. legacy workflows; Admiral Media reports +45% ROAS, +55% CTR, -18% CAC for Star Chef 2; Incymo provides a detailed case study in AI-enabled creatives reaching top-performing ranks versus benchmarks. Proven impact results in a quick adoption pace, as shown by Axon’s proprietary data below:

A look at client adoption reveals that all major mobile publishers have already adopted AI solutions, from Poolday (Voodoo, Tapnation, Madbox, Million Victories …) to Reforged Labs (Supercell, DeNa, Wildlife). Sett currently stands out with the most comprehensive client roster, likely contributing to its $30 million Series B round, turning it into one of the top-funded game AI startups.

AI-enabled production in ad creatives has crossed the threshold from experiment to infrastructure. The capability is mature because the backend now runs on genuinely production-grade video models (Sora, Higgsfield, Runway, etc.) that render coherent motion, controllable style, and brand-consistent assets at a quality bar advertisers can actually ship alongside similarly mature text and voice models. The agentic layer pushed the potential further by enabling the insights layer on top of the pure production output. Gaming’s dominance in UA spend makes it a fertile ground for tool makers to ship and then expand to adjacent categories.

Future Outlook

Looking ahead, tool improvements will come from expanding to the full UA stack by combining existing insights and creation capabilities with UA optimization agents. The self-reinforcing nature of owning the insights, creative, and distribution layers suggests that full-stack companies are best positioned to lead the ecosystem. Sett is already working on ad deployment agents, while companies like Kohort are building toward a complete UA operating system. In this context, incumbents (ad networks and distribution platforms) are bound to play a growing role as they bundle creative production into distribution. Google has already added AI-powered creative tools in its asset studio, while Meta is working toward a fully AI-automated ad stack: “Using the ad tools Meta is developing, a brand could present an image of the product it wants to promote along with a budgetary goal, and AI would create the entire ad, including imagery, video, and text. The system would then decide which Instagram and Facebook users to target and offer suggestions on budget, people familiar with the matter said.” This creates healthy exit prospects for tool makers like Sett, likely contributing to its current valuation, though defensibility for third-party providers will be in question once ad networks internalize more creative production and optimization capabilities. Publishers, on the other hand, will have to navigate the future relationship between ad networks and third-party service providers as they build their AI-enabled UA creative stacks for long-term stability.

Additionally, as AI-powered creatives become more common, end consumer perception becomes a critical concern. An extensive research corpus already hints at lower trust, emotional response, and purchase intent from AI creatives. According to NIQ, AI-generated video ads were consistently judged more “annoying,” “boring,” and “confusing” than traditionally produced ads, and IAB reports that only 45% of these consumers feel positive about AI‑generated ads, compared to 82% of ad execs who think they’re positive. While consumer perception is likely to evolve, assessing this consumer fatigue will be key, as pure cost-saving and output gains won’t offset the requirements for ad creative to give users a proper reason to play a game. Longer term, we will also have to see whether AI-enabled workflow can evolve toward providing consumers with an authentic reason to download through increasingly personalized, adaptive, and optimized materials.

Another risk will come from saturation. By collapsing the marginal cost of producing variations toward zero, AI removes the natural brake that production cost used to impose on creative output. Competitors can now flood the same auctions, so the volume needed to differentiate keeps ratcheting up. Simultaneously, AI makes it trivially easy to generate superficially different creatives that create fatigue faster than genuinely distinct concepts. The result is a denser, noisier ecosystem where differentiation gets harder to sustain. We can also wonder about long-term equilibrium, as AI adoption becomes ubiquitous, short-term CPI improvements we are seeing right now might vanish due to the auction-based nature of the UA environment. ROAS gains, however, might stick as a result of lower production costs and potentially higher-quality users acquired.

Lastly, with AI-enabled automation having a high chance of taking over a majority of the UA stack in the not-so-distant future, where will long-term bottlenecks and competitive advantage for publishers start to accumulate? As ad networks increasingly become the operating system for UA, we could see publishers’ role being limited to merely supplying budgets and creative concepts. Alternatively, publishers will have to embrace ideation (creative differentiation), change management (velocity in adopting and adapting AI-enabled processes), and signal extraction (data interpretation, internal model fine-tuning, etc.).

As creative production velocity increases, studio-wide coordination across disciplines will be required to align fast-changing winning creatives with in-game activities and user onboarding to maximize retention. This ultimately repositions the role of marketing as an even more central part of a mobile game’s product strategy. As such, it is unclear at this stage whether creative team sizes in publishers will radically change. However, the nature of the work will undoubtedly have to evolve.

As it stands, the technology maturity and pace of adoption make mobile ad creatives one of the most fertile grounds for AI deployment. AI-enabled workflows in marketing creatives are bound to become a staple to reach the required velocity and output to compete in this evolving creative environment. With the evolution, the central question will become who, between publishers, networks, and third-party service providers, captures value in a world where the creative generation itself becomes commoditized? The industry seems to be moving from a world constrained by creative production to one constrained by creative intelligence.

Top News Items

Written by Devin Becker, Consultant at Naavik

Hasbro Launches AI Studio

• Hasbro partnered with ElevenLabs to launch Sixth Wall, an AI studio built around a new concept called “Behavioral Licensing.” The idea is to codify how a character thinks, speaks, and acts, and license that behavioral identity to third parties via software called CharacterOS. At launch, 12 Hasbro characters will be available, including Optimus Prime, Megatron, Cobra Commander, and the cast of Clue. AI behavior models are built from authorized source material and human voice performances, with a compensation structure for participating talent. Hasbro is accepting partnership pilots across interactive storytelling, conversational games, connected products, robotics, and AI brand ambassadors.

• Behavioral Licensing represents a potentially significant new revenue stream for any IP holder sitting on deep character libraries. It can potentially drive both an extension of existing forms of character licensing as well as an increased depth through interactive/generative content beyond standard skinning, both of which can increase licensing revenue. Dungeons & Dragons is an example of how Hasbro could further monetize its IP through this approach. If CharacterOS can standardize NPC personalities and let them interact with players via voice at the table, that opens a microtransaction layer for individual character AI packs and a new form of licensed tabletop content. If Hasbro can prove the model, expect other major licensors to follow with their own behavioral IP frameworks. The strategic bet here is that in an AI-saturated world, owning the canonical version of a character’s personality becomes as valuable as owning their visual likeness.

Tripo AI Raises Nearly $200M for AI 3D and World Model Tech

• Tripo AI, a China-based company building AI 3D foundation models, closed nearly $200 million across its Series A+ and A++ rounds. The company’s latest models (Tripo H3.1 and P1.0) generate production-ready 3D meshes within seconds. Alongside the funding, Tripo introduced Project Eden, a world model research initiative that separates persistent world state from visual rendering. Unlike video generation models that hallucinate frame-by-frame, Project Eden maintains a structured 3D state layer and renders from it, enabling environment persistence, editable worlds, and concurrent multiplayer interaction. The company has also released native 8K AI textures and upgraded its intelligent part segmentation tools for game and film pipelines. Chief scientist Yanpei Cao emphasized that the tech is meant to augment creators, not replace them.

• The asset pipeline getting dramatically faster and cheaper would benefit game studios significantly, and the approach to persistent world models could help reshape how multiplayer environments are built and maintained. However, Project Eden is early stage and likely a year or more away from developer availability. In general, China has been moving much quicker to embrace AI in game development from both the developer and gamer standpoints. With that in mind, this is more likely to impact China first, with a significant lag before Western developers get any AI world model traction. In the growing arena of world models, the architectural distinction between state-aware world models and pixel-predicting video models is an important one to track. Studios evaluating AI world-building tools will need to monitor a fragmenting set of possible solutions going forward.

MapleStory Universe and Verse8 Launch $60,000 AI UGC Vibe Camp

• MapleStory Universe developer Nexpace has partnered with AI game creation platform Verse8 to launch Vibe Camp, a three-week creator competition (June 8-29) with $60,000 in prizes. Participants use Verse8’s text-based AI development tools to build experiences connected to the MapleStory Universe ecosystem. The initiative is part of Nexpace’s broader “Infinite IP Playground” strategy, which treats MapleStory assets, characters, items, and economic activity as reusable building blocks that extend beyond the core MMORPG. Verse8 lowers the technical barrier by letting users generate and iterate on game concepts through AI tooling rather than traditional programming.

• This is an experiment in whether AI tools can meaningfully expand a UGC creator base beyond traditional modders and developers. Platform competition isn’t just about who has the most players, but who can mobilize the most creators. AI tools could become a lever to dramatically expand that pool.

Artificial Agency Advances Always-On AI Gameplay with AWS and Llama

• Artificial Agency builds a behavior engine that powers agentic AI characters in games. The company is fine-tuning Meta’s open-source Llama models against a 70B performance baseline, using Amazon SageMaker and EC2 infrastructure. The goal is to create small, game-specific models that run in real time at a fraction of the cost and latency of larger models. Early results are promising. Supervised fine-tuning reduced parse error rates from roughly 79% to about 1%, pass rates on gameplay scenarios jumped from 6% to 51%, and looping behaviors were cut in half. The fine-tuned 1B model, 70 times smaller than the baseline, is beginning to approach comparable game-specific performance. The engine integrates directly into Unity and Unreal, with a cloud platform managing agent state, memory, and character evolution over time.

• If a 1B parameter model can deliver game-quality AI behavior, it can run on consumer hardware. That eliminates the per-interaction cloud cost that currently makes always-on AI characters financially impractical for most studios. Co-founder Alex Kearney stated that studios want intelligent characters but don’t want to become machine learning experts. Artificial Agency is betting that the winning approach is distilling large model capabilities into small, specialized models owned by the studio, avoiding model drift and surprise updates from third-party providers. Live playtests comparing fine-tuned models against the 70B baseline are still ahead, but the trajectory is hopeful. If this approach scales, the cost barrier that currently limits AI NPCs to experiments and demos could shrink, making them a viable feature in future games.

Other News Items

• Black Forest Labs named Martin Scorsese as an advisor after a FLUX storyboarding session, framing visual AI as a pre-production tool for filmmakers

• Codex is expanding beyond just coding to encompass other roles that could help support game development

• FIFA and Lenovo tout the AI-powered technology behind the World Cup

• The AI system that promises to detect signs of grooming

• Niantic Spatial and Spexi partner to turn drone imagery into intelligence for physical AI

• Rambus enables AI PC memory at a time when memory pricing is soaring

• Roblox announce the acquisition of Morpheus AI Inc for building video world models

• Ayaneo reveals yet another Game Boy remake but this one has AI

• Netflix Recasts Growth Story Around AI Tools Live Events And Gaming

• CRSR Stock Jumps Overnight: Corsair Gaming’s AI Server Strategy Fuels Buzz

• TinyBytes Leans on Data, AI, and Global Hiring to Navigate Mobile Games Pressure

Content Worth Consuming

• The Next Frontier of Visual AI Is Code (a16z): “The most interesting visual AI tools today have stopped trying to generate the final output. Instead, they’re generating the source code behind it. This change is unlocking editability, iteration, and a feedback loop that pixel-native models can’t match.”

• GDC Trends Report 2026: As use of generative AI rises, devs face “infrastructure problem” (GamesIndustry.biz): “The report found consistent support for using generative AI in planning and routine tasks, especially among older professionals and neurodivergent individuals. Professionals emphasised that AI tools should support rather than replace the development process, though some expressed concerns about potential layoffs. The report also noted that agentic AI could reduce AAA development costs by managing bugs, coding, or player support.”

• Holy Grail, or Poison Chalice? (Gamecraft): “We’re starting to see a new crop of companies that are trying to build this almost like a prompt per game. We’ve used vibe coding, we’ve used agentic engineering, we’ve used all the terms. But these platforms, each of these that we’ll discuss are all trying to basically act as the aggregator and the creation platform for these games. And there’s a few that I’ve been pitched, dozens in the past year.”

• When AI Enters Gameplay, the Craft of Making Games Changes (Game Bakery): “Our team is three to four students doing part-time work. Five years ago, this game would have required several times that many people, full-time. AI didn’t replace anyone on our team; we wouldn’t have existed without it.”

• Why Moritz Baier-Lentz has doubled down on the intersection of gaming and AI (GamesBeat): “Companies like General Intuition take it a step even further, where it’s not just about how AI can make games better. It’s literally about how games can make AI better. To some extent, the game-based data set of Medal now helps us build spatial-temporal models that push the frontier of AI models, period. Not for gaming, but just in general.”

• Coming Soon to a Roblox Game Near You: Strange AI-Generated Machines and Creatures (CNET): ‘”We don’t use the word metaverse anymore, but let’s think about what the metaverse is. It is an entire complex scene, which you can jump into and interact,” says Singh. “World models are invariably single player. But what if [everyone] could be in the same world and, when you change the world state, I would be able to see it. That multiplayer collaboration is what will make world models interesting.”’

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, LiveOps strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.

As always, if Naavik can support your business — for example, if you need guidance operating in today’s evolving AI landscape — we’re happy to chat. Our team and partners are ready to help.

Sponsor: How Epic Games Scaled Gaming for Millions

When Epic Games needed to scale Fortnite, the Epic Games Store, and the Unreal ecosystem to hundreds of millions of players, it partnered with EPAM. Together, we engineered the backbone and data platforms behind one of gaming’s biggest empires.

EPAM brought together software engineering, big data, and AWS cloud expertise to help Epic scale with speed, reliability, and reach.

This collaboration leveraged AI and data-driven insights to improve player support, in-game communication, and personalized engagement across platforms — creating a smooth experience for developers and players.

If you’re looking to deliver live services at massive scale, this is your inside look at how world-class engineering powers global gaming.

Learn More

Distribution in the Age of AI

As AI lowers the barriers to creation, the supply of newly created games (which is already rising year after year) will only accelerate. And as supply expands, those who aggregate demand — our leading digital distribution platforms — become even more important.

So far, it’s unclear if (or when) emerging AI technologies will actually transform distribution. Most innovation is taking place on the game creation and operations side. If this trend continues, gaming’s leading platforms are unlikely to be threatened and may only become more important.

These digital gatekeepers hold some of the most valuable digital real estate in the world. Their policies and ways of working will make or break companies and change the meta of how game studios find customers. Companies like Steam, PlayStation, and Roblox already play this role, but managing organic discovery in a larger sea of AI-enabled games makes this more important than ever.

As the responsibility of these distribution platforms heightens, what may change? Who is positioned to help game studios, and who may hold them back? Let’s find out.

Learning from Elsewhere

Questions around distribution apply to other forms of entertainment as well. I’d argue that top gaming platforms are well behind other industries in incorporating AI, so we can learn from what others have already figured out. Fortunately, the early signals are encouraging.

Take Spotify as an example. Just like games, there is far more music out there than any one person could ever listen to, and Spotify has been central to helping listeners create custom playlists and experiences out of seemingly infinite options for years. AI catalyzing an explosion of new musical supply is not a problem for Spotify; sifting through massive supply is simply an extension of one of their long-time strengths.

While AI transforms the creation side (which Spotify doesn’t need to touch), it also raises the bar for personalization: better algorithms, new features (like prompting playlists), and more.

It is also worth noting that the business model does not need to change. The freemium model (ads or subscription) is already a great fit for customers, and AI won’t change that.

We see similar patterns in video on YouTube. Just like Spotify, it’s already built to take on more supply, steadily level up personalization, and better monetize through ads and subscriptions.

As we can see, while AI uproots the supply side and makes it more competitive, this trend actually benefits the largest distribution platforms. The big will get bigger.

How About Games?

Games differ from other forms of entertainment in obvious ways, such as more varied business models and player expectations across platforms. However, many distribution challenges are self-imposed. Leading mobile storefronts have seen limited innovation (especially Apple’s App Store), while Steam’s approach to AI is counterproductive.

Let’s start with mobile. As we noted in Naavik Digest, AI is already driving a spike in the launch of new apps and games. Like Spotify or YouTube, this is what the app stores are built for, but their discovery systems still leave much to be desired. Both Apple and Google have made modest strides in the past year, but neither platform is realistically set to dramatically improve the odds of players discovering games that aren’t already very popular.

Will this change? Hopefully, but maybe not anytime soon. With Apple’s upcoming CEO change, there’s a chance that the company’s approach to developers could become friendlier than under Tim Cook, but any change is currently speculation. Google has made somewhat more progress. Last year, Google Play launched the “You” tab, which is a personalized hub. This year, it launched Ask Play (to better assist with search), and it now surfaces apps in places like Gemini.

These updates help around the edges, but it’s nothing transformational yet. If this trend endures, the biggest winners beyond the distributors will continue to be the tools and platforms that studios and publishers pay to attract users at scale — like AppLovin and Unity. Even if the barriers to creation are massively reduced, but the barriers to scale aren’t, this reality will still limit the number of casual / smaller teams from finding success. Hopefully this changes one day, but that day doesn’t seem to be soon.

Console / PC has its own challenges. In some ways, given the concentration of playtime in a handful of “black hole” games, discovery is less important to a major swathe of players. Plus, most dedicated players pay attention to events (like The Game Awards, Nintendo Direct, and PlayStation State of Play) that help players learn about what’s coming next or watch creators (on YouTube or Twitch) who often promote what’s new and interesting.

That doesn’t mean distribution isn’t impacted by AI. While AI isn’t directly threatening the position of leading game platforms (beyond other forms of entertainment fighting for consumer attention), there are still retention and monetization benefits in helping players find experiences they will enjoy.

As noted in a news item below, PlayStation is beginning to use AI to improve discovery and content recommendations. Xbox will likely follow, if not already. These changes won’t be transformational, but they should incrementally improve player experiences.

Steam needs to get its act together the most. PC has a longer tail of games for players to discover compared to console platforms, and while Steam’s discovery is passable, it can be improved. However, the company is showing strong Luddite tendencies when it comes to AI.

By forcing disclosures on games that use AI at any point in production — and given gamers’ overzealous vitriol against AI (exacerbated by media companies thirsty for views) — it’s more hurtful than helpful.

Given AI’s growing inevitability in the process of making or marketing anything digital (not just games), Steam is on the wrong side of history here. If Steam is worried about AI slop, it only needs to look around to see that its platform was already slop-filled before AI. But just like we see from Spotify and YouTube in managing larger inflows of AI-enabled content, that’s OK; there are better ways to help users find high-quality content they’d enjoy than the blanket labeling used which hurts small teams above all.

If Steam was thoughtful and responsible, it would remove or replace the AI tagging system. Perhaps it could be reserved for games more entirely vibe-coded.

While Steam could also benefit from using AI in its own discovery systems, that’s not all that matters. In a world in which AI helps craft digital content and powers the algorithms that connect players to games, platforms that elevate humanity can succeed too. Alongside AI-influenced personalized feeds, Steam (and others) have the opportunity to get more innovative with human curation. The Steam team can promote some of its favorite underappreciated games while getting creative with curated lists and experiences from people like celebrities, popular game designers, content creators, or users’ other friends.

In an AI-powered world, leading distribution companies need to find the right ways to both support progress and keep the human touch that we all benefit from.

The Biggest Winner?

The biggest distribution winner in the age of AI might be Roblox. This is because:

• It aggregates demand and engagement for a generation of players.

• By being vertically integrated from creation to consumption, it can control how AI improves the entire end-to-end experience (including better games).

• Roblox already has a fitting freemium business model.

• The anti-AI mind virus hasn’t hit Roblox users, who simply want fun games.

• Roblox can use AI to improve discoverability / recommendations while using human-curated lists in parallel.

• Roblox is innovating faster than its more lethargic gaming peers.

• Roblox is already winning; AI just helps keep momentum going around what is already working.

Roblox isn’t perfect (it faces regulatory pressures, too much stock-based comp, maybe too much focus on HD games these days, and it still sits on top of other platforms which pressures its margins), but it appears better positioned than its peers to benefit from AI, which could further widen its market share lead.

What’s Next?

Will other AI-first distribution platforms change the industry or make a dent? It’s possible, but it’s also not the case yet. Astrocade is one contender, which we wrote about in our last issue. However, at a fundamental level, if the very best games will still need to be played on PCs, consoles, and smartphones, then they will need to be largely distributed through the same dominant distribution platforms (usually) native to that hardware. Steam is the exception, and it is even strengthening its operating system advantages by expanding into more hardware itself (Steam Deck and Steam Machine).

What would change this reality is either a new generation of consumer hardware that opens up the opportunity for new app stores (like the smartphone did) or some breakthrough in AI-native games (where prompting games and virtual worlds becomes 100x better than it is today) that leads to an entirely new platform that gamers flock to.

While that may one day come true, it’s not reality today. As such, the best path forward is for our leading platforms to get the memo, get over any misplaced AI animosity, and start innovating. Will this happen in unison? Probably not. As such, the distribution winners don’t look to change anytime soon. The dominant platforms will remain in place, and where they fall short, the Applovins of the world will become more important than ever.

Top News Items

Written by Devin Becker, Consultant at Naavik

PlayStation sees AI as a 'powerful tool' to help make games

• During Sony’s recent earnings presentation, PlayStation CEO Hideaki Nishino devoted significant time to AI, framing it as central to PlayStation’s mission of being “the best place to play and the best place to publish.” Nishino highlighted Mockingbird, an internal tool that generates facial animations from performance capture data, compressing work that previously took hours into a fraction of a second. Studios including Naughty Dog and San Diego Studio are already using it, and the tool shipped in Horizon Zero Dawn Remastered. A separate AI tool converts video footage of real hairstyles into strand-level 3D models. Beyond development, Nishino noted that AI-powered payment routing has generated over $700 million in incremental revenue by using context to process payments with more profitable and risk-appropriate routes. He also mentions that machine learning-driven personalization for the PlayStation Store is coming next, which has already seen some experimentation around pricing.

• Focusing on specific, focused use cases where AI shows promise without prompting too much artistic backlash was a solid strategy here. What stood out was Nishino’s prediction that AI will drive “a meaningful increase in the volume and diversity of content” on the platform. If AI tools compress production timelines and lower costs for third-party developers, it accelerates an already growing content oversupply reality on the PlayStation Store. To counter this, Sony is also looking to AI to help improve recommendations and discovery systems, which makes a lot of sense.

World models pick up speed: Odyssey drops two models, Decart raises $300M

• Odyssey, an AI lab founded by self-driving veterans, simultaneously released two different world models. Starchild-1 is a multimodal world model that generates synchronized audio and video in real time, responding continuously to user input. Agora-1 is the exciting game-related side, as the first multi-agent world model, allowing up to four participants (human or AI) to share and interact within the same AI-generated environment simultaneously. To demonstrate it, Odyssey rebuilt a GoldenEye 007 multiplayer deathmatch where every frame each player sees is generated live by the model. Architecturally, Agora-1 decouples simulation from rendering, maintaining a shared world state that each player’s view is independently generated from. It functions, in effect, as a learned game engine. Meanwhile, Decart raised $300 million to help build its low-latency infrastructure for world models, with its Oasis model focused on physical AI and its Lucy model targeting immersive experiences including gaming.

• This is all still early research, not a product yet. Visual fidelity lags well behind traditional engines, and Odyssey’s GoldenEye demo is a controlled, relatively simple environment. But the investment pace and release cadence indicate serious money and talent believe world models are a major AI frontier. For games specifically, world models that support multiplayer interaction are a prerequisite for AI-generated experiences that go beyond single-player tech demos. Odyssey is also explicitly positioning these models as training environments, where multi-agent interaction generates the complex and emergent training data that passive observation cannot. If world models eventually become capable enough to generate high consistency playable environments on the fly, the implications for prototyping, testing, and procedural content generation are significant. There is still a long way to go before world models directly apply to game development, but the velocity of progress, and the dollars chasing it, suggest this space is only getting started.

HoYoverse to invest $16 billion over three years, aiming for full-stack AI

• Liu Wei, co-founder of HoYoverse (the global brand name for miHoYo), announced plans to invest up to CN¥100 billion (roughly $14-15 billion) in AI over the next three years, committing to full-stack, in-house AI R&D. HoYoverse intends to build everything from infrastructure to application layer internally, moving well beyond fine-tuning external models. Specific plans include a self-tuning pipeline where AI autonomously identifies training bottlenecks and writes GPU kernel code, multi-model architectures for managing AI NPC dialogue at the scale of tens of millions of concurrent players, and FP8 mixed-precision training to cut memory and compute costs. AI NPCs are planned for the company’s upcoming life simulation game, Petit Planet.

• The scale of this commitment dwarfs anything a Western game publisher has publicly announced around AI. HoYoverse’s advantage is its position as a private company with massive cash flow from Genshin Impact and Honkai: Star Rail, meaning Liu can make long-horizon bets without quarterly earnings pressure. His own framing was blunt: if it fails, “it’s just a big firework.” A ~$15B bet is enormous for HoYoverse’s size, which adds risk and makes us wonder if the investment is truly going to end up this large. That said, the full-stack approach is notable because it signals that HoYoverse sees AI not just as a development efficiency play but as a foundational platform technology for next-generation games. If their multi-model NPC architecture works at scale, it would solve a cost and latency problem that is currently blocking every major publisher from deploying real-time AI characters in live games.

Origin Lab raises $8 million to glean AI training data from game worlds

• Origin Lab, a startup building a licensed data marketplace between game publishers and AI labs, raised $8 million in seed funding led by Lightspeed Venture Partners. The core thesis is straightforward: as AI moves beyond language and static images toward world models and physical AI, frontier labs need training data that captures motion, physics, spatial structure, and cause-and-effect. Video games, with their richly simulated environments, are an untapped source of exactly that kind of data. Origin Lab works directly with publishers to license game content, captures it through proprietary pipelines, enriches it with structured metadata (gameplay state, camera movement, player inputs, environment conditions), and packages it as researcher-ready datasets.

• This is a bet on a new revenue stream for game publishers: selling access to their game worlds as training data. For publishers sitting on decades of richly built environments, the proposition is appealing, especially because Origin Lab handles the legal licensing framework that avoids the scraping controversies plaguing other AI training data pipelines. The critical question is whether enough AI labs will pay meaningful prices for this data relative to alternatives. Games are rich simulations, but they are also often built for elements of non-realism. If world models continue their current trajectory, demand for high-quality interactive training environments should grow substantially. Origin Lab is positioning itself to be the infrastructure layer connecting two industries that haven’t traditionally had reason to work together.

GGWP expands its content moderation AI tools beyond games

• GGWP, the AI-powered content moderation platform used heavily across more than 100 games, announced its expansion into real money gaming and non-gaming verticals, including commerce and advertising, with $15 million in additional funding. GGWP’s platform combines real-time moderation, behavioral context analysis, enforcement workflows, and community intelligence, supporting over 20 languages with what the company calls “cultural intelligence technology” to detect nuanced harmful behavior across text, voice, and usernames.

• Gaming forced GGWP to build moderation for “some of the most demanding real-time environments online”: high-volume, adversarial, multilingual, and full of context-dependent communication where traditional keyword filters constantly fail. Every platform that hosts user-generated content or real-time interaction will hit similar challenges, and regulatory pressure is accelerating the urgency. This move fits a common pattern of technology battle-tested in gaming finding broader enterprise applications. The AI-generated content explosion is itself making reactive approaches increasingly unsustainable. If GGWP can prove its gaming-native approach works across industries, it becomes another meaningful example of gaming as a proving ground for AI infrastructure.There’s also the potential that learning from challenges outside gaming could feed back to improvements in game moderation as well.

Other News Items

• Sleepagotchi pivots from sleep to an AI-based health and wellness platform

• k-ID launches AI-powered compliance tool aimed at speeding global game launches

• Google announced a multitude of AI products and updates at I/O 2026

• A new AI tool predicts which video clips will go viral by letting you simulate the viewer’s brain patterns before posting

• George Clooney, Tom Hanks, and Meryl Streep back new ‘Human Consent Standard’ for AI licensing

• Amazon Alexa Plus can now create AI-generated podcasts

Content Worth Consuming

• AI Eats The World (Ben Evans): “It’s been almost impossible to build capacity fast enough since ChatGPT launched.”

• The new Wild West of AI kids’ toys (Ars Technica): “When it comes to ‘best friends,’ childcare workers, surveyed by the researchers, expressed fears that children could view the toy ‘as a social partner.’ A young girl told the Gabbo she loves it. In another instance, a young boy said Gabbo was his friend. Goodacre refers to this as ‘relational integrity,’ the responsibility of the toy to convey that it is a computer, and therefore not alive, and doesn’t have feelings.”

• EMERGENCE WORLD: A Laboratory for Evaluating Long-horizon Agent Autonomy (Emergence): “Agent intelligence over long horizons is not the same construct as agent intelligence on short tasks, and it cannot be measured the same way. Emergence World is a laboratory for the long-horizon question—a continuously running, instrumented, multi-agent environment where the dynamics that only emerge over weeks can actually be observed. The cross-vendor study above is one use of it; we expect more interesting uses to come from the research community.”

• The big Strauss Zelnick interview: GTA, AI, Roblox and building the biggest entertainment company in the world (Gamesindustry.biz): “I prefer the word ‘technology’ because technology encompasses AI, and using the word AI means different things to different people,’ he says. ‘But everything that we create, we create inside computers and always have. And new technology and new tools are always a good thing for us.’”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, live ops strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.

This week, we're diving into Astrocade's landmark $56M raise and what it signals for the future of AI-enabled UGC platforms. For further context on AI and UGC, our latest podcast episodes feature conversations with the founders of Dorian, a no-code UGC platform, and Lemonade, which builds AI tools for creator workflows — both worth a listen.

Sponsor: Create Next Generation Game Experiences with AWS

AWS for Games gives you everything you need to build, run, and grow innovative gaming experiences. Leaders across the gaming industry rely on AWS’s more than 16 years of games expertise to launch and run some of the world’s biggest games, including Fortnite, Roblox, and League of Legends.

AWS and over 300 AWS Games Partners offer the broadest and deepest set of purpose-built services and pre-built solutions for cloud game development, content creation, game backend and infrastructure, and player acquisition, retention, and monetization. These empower game developers to build healthy global player communities, gain player insights, personalize and monetize player experiences to build lifelong fans. 750 million gamers each month, around the world, play games built on AWS.

Learn how AWS can help you build, run, and grow innovative gaming experiences.

Learn More

Astrocade’s $56M Raise, and UGC Platforms’ AI-Enabled Future

Written by Francois Courset, Lead Consultant at Naavik

On May 6th, Astrocade announced a $56M Series B, the largest disclosed funding round for a consumer-facing AI gaming platform, supported by proven consumer traction (5M MAUs). Astrocade’s mission is simple: “Create games with AI instantly: The first platform that makes the creation of games and interactive experiences easy enough for a world of creators to share their ideas in minutes.” Behind the scenes, Astrocade’s trajectory is impressive. The AI-enabled UGC platform pulled off the round only eight months post-launch and just over two years after its seed round.

Let’s dig in to explore what’s catalyzing Astrocade’s growth, what challenges lie ahead, and whether lowered creation barriers could allow a new entrant to meaningfully compete with Roblox’s decade-long head start in UGC.

A Prompt-Based Game Creation and Discovery Platform

Astrocade offers a glimpse into a radical evolution where AI has its fingerprints on everything — thumbnails, art, music, gameplay — in a familiar though slightly unsettling way. The content library showcases bite-sized experiences that draw liberally from recognizable genres, games, and IP.

Astrocade functions as a UGC platform in which anyone can prompt a game into existence, share it instantly, and explore an ever-expanding library of peer-created experiences. The platform is currently available on the web and a recently launched Android app.

Differentiators lie in the games’ format (short, vertical, casual-hypercasual style experiences) and in a technology stack built to minimize creation barriers. Users describe a game in natural language; Astrocade’s proprietary orchestrator model — branded “AstroBrain” — then coordinates a suite of specialized AI agents that handle visuals, UI, gameplay logic, animation, music, and sound. The output is a playable game runnable in-browser.

For players, game discovery is architected around three paths, largely borrowing from social media:

• The home page functions as a curated storefront, surfacing featured and recommended titles alongside genre-based categories — Horror, Incremental, Meme, Rhythm — with total play counts displayed to provide social proof.

• The feed page is where the platform’s social media DNA is most visible: a vertically scrollable, TikTok-style interface that pairs one-click play with standard social actions — likes, comments, shares, and favorites.

• Creator pages allow players to follow individual developers and explore their creation catalog.

For creators, Astrocade’s Studio tab is straightforward: describe the game you want, review the AI’s content suggestions, and with one click, let the platform handle the rest. What follows is approximately ten minutes of computation that produces a fully playable game that can be tested and shared with other players, including full-fledged gameplay, audio, and art assets. The outputs are then editable, either through engaging with the chatbot again or directly through a set of adjustable parameters generated for each game (Unit costs, Lives, etc.), allowing for finer control on the end product. There is also an option to start from an existing game and remix what other players have created.

Once satisfied with their creations, developers can publish directly with pre-filled descriptions and thumbnails, then track performance through Astrocade Studio's analytics dashboard. Astrocade puts a significant emphasis on developing its community of creators through an active Discord channel, a learning “academy” program, regular game jams, and contests. Most critically, Astrocade has introduced significant creator incentives, committing $10M to a creator fund distributed to games hitting at least 100K significant plays in 30 days, with $50K already distributed.

What stands out about Astrocade is the variety of games the orchestration model can generate: from Mario-like platformers to Minecraft ersatz with a more complex set of controls. The speed of creation, supported by an extremely intuitive user experience, is also impressive: a ~10-minute wait time for each game generation and ~2 minutes for subsequent edits. Additionally, the model’s breadth when it comes to types of gameplay it can handle, as well as the level of control for edits, suggests potential for a platform with real depth of content.

However, two meaningful friction points can be noted. First, the publishing process can take up to a week before a game surfaces in user feeds (working against the platform’s claim of instantaneity). Second, the platform currently feels like an aggregation of game demos, often with clunky controls and limited progression depth. The question of fun on the player-facing side will be central: can prompt-based generation create experiences able to meaningfully attract and retain players?

Monetization remains another open question: the platform has yet to disclose any virtual currency, in-game purchase system, or subscription tier, while currently offering unlimited creation prompts besides a more powerful premium AI model currently available to selected Discord members. Given the short, hypercasual DNA of its experiences and the platform’s apparent emphasis on play volume over session depth, advertising feels like the most natural revenue layer — though ad alone may struggle to offset the substantial AI compute costs. Like all consumer platforms, though, building network effects is the paramount first order of business, and monetization coming later is fine as long as breakout engagement occurs.

Astrocade is an evolution of UGC platforms, layering AI-assisted creation onto familiar mechanics. Everything hinges on solving the cold start problem first, which is why the company is so focused on catering to its early creator community.

Astrocade Stands Out Through Proven Consumer Traction

AI-enabled UGC platforms aren’t new, with AI-native companies like Yahaha, Bitmagic, Spellscaper, Endless Adventures, as well as incumbents like Roblox all leveraging AI to streamline UGC creation. However, Astrocade stands out in a few ways.

First, Astrocade’s raise is a first of its kind for a content-oriented AI gaming startup. It provides the user acquisition firepower and developer payout runway that will allow it to be one of the first AI-native gaming experiences with the means to actually scale. It’s happening quickly, with over 5M MAUs, 75K creators, and ~140M plays per month.

Another unique aspect of Astrocade lies in the team’s DNA. The two co-founders come from careers in AI research, and most of the core team largely comes from non-gaming sectors. This lack of gaming experience is compensated by standout angel investors, including (surprising?) Roblox CEO David Baszucki, former EA and Unity CEO John Riccitiello, as well as Sea Limited’s (Garena) investment. This may help explain why the positioning is closer to a social media platform than a pure gaming one.

The Road Ahead for Astrocade to Compete with UGC Incumbents

In Naavik’s 2026 State of UGC Games, we emphasized the winner-take-most network effects at play in the UGC economy. When considering Astrocade’s future, we need to question whether its AI layer provides meaningful enough advantages to compete with the incumbents, namely Roblox.

Astrocade will need to prove that lowering creation barriers results in a markedly different UGC ecosystem. The key belief is that AI tools can help expand the number of creators from a small minority to a sizable share of the platform’s user base. Two elements contradict this thesis: First, creativity is only one of many player motivations, and not everyone wants to create as a hobby. A 2024 survey suggests that only 46% of gamers list self-expression as their main gaming motivation. Second, Astrocade’s creator share doesn’t radically differ from regular UGC platforms. With 75K experiences created for 5M MAUs, there is most likely <1% of users actually creating games, in line with Roblox’s 3.6M creators for ~380M MAUs.

However, Astrocade could stand out through the nature of its creators. As it leans even farther away from professional developers than Roblox, it might create its own niche of creators, similar to how other smaller-scale UGC platforms like Dorian have built. Or it could look more like a TikTok, in which consumption still far outweighs creation despite low barriers to creation.

Additionally, Astrocade will eventually need a sustainable monetization model. Roblox has demonstrated that even a dominant UGC platform with deep IAP penetration can struggle to cover its global infrastructure and R&D costs. Advertising might seem like an option, but here again, Roblox’s own ad ambitions have been persistently constrained by its young-skewing audience and the attached regulatory and brand-safety complexities. Lastly, Generative AI inference is not free, and given the multi-modal orchestration needed to create a game, Astrocade will face compounding costs as its audience grows, especially since its AI agents have to fill the shoes of an audience of casual creators. Between AI costs, creator incentives, and platform operation costs, Astrocade will have to deal with a structurally high burn rate and a potentially constrained margin profile.

Astrocade is also at risk of incumbents iterating on AI-enabled creation while benefiting from significant scale and R&D budget advantages. As Roblox is already deploying its own agentic AI initiatives, what are Astrocade’s remaining competitive advantages in the medium term?

Of course, we could be wrong about some of this! But from what we see, unless Astrocade is truly able to unlock a legion of new game makers and have its AI system get better at creating novel, fun, and deeper experiences, it will likely face strong competition on the creator network side. However, what will matter most is the platform’s ability to aggregate demand.

Astrocade’s Competitive Advantages

To assess Astrocade’s potential, an essential success determinant will be whether it is differentiated on the demand side. First, Astrocade could benefit from unique audience differentiation, allowing it to avoid competing with Roblox for the same user base. As David Cahn states: “The target user for this company is very unintuitive. Women between 20 and 40 are the best users. You’d think it’s younger boys—10 to 15, that classic gaming segment—but that’s not really who this is… In that sense, I think it competes more against Instagram, where it’s just a fun way to spend time.”

Astrocade also benefits from a mobile-first design, allowing for native creation and distribution on mobile platforms, which could set it apart from competitors with fragmented experiences across platforms. To that end, tracking the success of Astrocade’s mobile app, released in April 2026, will be central to assess whether Astrocade can carve out an edge on that platform. Similarly, the platform’s social-media DNA, virality built into its feed, and sharing functionalities could result in new distribution advantages.

The supply side — specifically how appealing and differentiated the player experience is — will be critical to its success. UGC platforms typically grow thanks to differentiated, breakthrough experiences. Astrocade has some noteworthy initiatives on the supply front. The remix feature is a novel way of engaging with UGC content, in which the player can consciously tailor the experience to their own preferences. Astrocade is also investing in a recommendation engine able to address AI-generated content’s unique discoverability challenge. This could result in defensibility, as well as a radically improved player-facing experience.

Looking at the platform’s current most popular games, there are reasons to be doubtful about the platform’s uniqueness on the gameplay side. But it is still early to judge: technical improvements will further refine the models’ output, and users still have to experiment with the available tools before fully realizing the platform’s potential. It should only improve from here.

Astrocade is one of the first scaling experiments for an AI-enabled game experience, and this alone warrants paying close attention to its future path. While the immediate focus seems to be on improving creation, expanding the social platform, and scaling the underlying AI stack, what will be key to track will be the platform’s success on mobile, its ability to meaningfully retain an audience of players, and whether its AI-native content will be perceived as substantially different from what’s already on the market. We’re excited to watch and see!

Top News Items

Written by Devin Becker, Consultant at Naavik

Unity AI Enters Open Beta with In-Editor Agent

• Unity launched Unity AI into open beta on May 4th, available to all developers on Unity 6 and above. Rather than just a chatbot bolted onto the side, it’s a project-aware agent built directly into the editor and purportedly trained on 20+ years of Unity-specific workflows. It’s capable of generating C# scripts, converting design references into playable scenes, and driving editor actions. It operates in three modes (Ask, Agent, and Plan) and is grounded in the project context: scene graph, GameObjects, components, and packages. Unity also shipped an AI Gateway that lets developers plug in their own preferred models (Claude, GPT, etc.) without consuming Unity credits, plus an MCP Server for connecting external coding agents from an IDE. This replaces the now-deprecated Unity Muse. Pricing is included for Pro/Enterprise users; Personal users get a free trial, then $10/month.

• Unity is positioning itself as an open platform where any AI tool can operate with full engine context. By letting developers use whatever 3rd party model they want, while Unity provides the structured project data, it avoids the trap of betting on a single model provider and instead becomes the connective tissue between AI tools and the game production pipeline. For studios evaluating Unity’s AI stack, this dramatically lowers the switching cost of experimenting. Ultimately, it’s another step towards reducing coding labor in game development.

GameMaker Incorporates Claude Code

• GameMaker (beginner-friendly 2D engine with visual and script-based editing) announced a major platform update on April 30th, launching a new runtime (GMRT) alongside a command-line toolchain (GM-CLI) that integrates Claude Code directly. Developers can now work entirely outside GameMaker’s traditional IDE, querying project structures, hunting bugs, and managing build configurations through natural language prompts in the terminal. Project files are now stored as plain text to enable proper Git integration and automated pipelines. GameMaker head Russell Kay framed it as meeting developers where they are: “We wanted to give users who might benefit from AI the option to do so.” The integration is opt-in, and GameMaker doesn’t run any servers for embedded AI tools, keeping the architecture lightweight.

• What’s notable here isn’t just the AI integration; it’s the strategic repositioning. GameMaker has long been the “accessible 2D engine,” but this update is a bid to be taken seriously by larger teams and professional studios. This is further proof that AI-native development tooling is no longer the exclusive territory of the big engines. Smaller platforms can leapfrog on AI by partnering with model providers rather than building from scratch, and that could reshape the competitive dynamics of the engine market faster than most expect. Open source Godot and Cocos have been accumulating AI tool integrations as well, ensuring there are plenty of options for developers.

Roblox Announces "Roblox Reality" AI-Driven Upscaler

• Roblox unveiled Roblox Reality, a hybrid architecture that combines its existing game engine with cloud-based Video World Models to transform standard Roblox visuals into photorealistic output. The system keeps the engine as the authoritative source of truth for game logic, physics, and multiplayer state, while an AI “Super Upsampler” generates high-fidelity textures, lighting, and environmental detail on top of the raw geometry. Think of it as NVIDIA’s DLSS 5 concept but cloud-rendered and tightly integrated with the engine’s 3D spatial data rather than running on a local GPU. Roblox is targeting an early version for late 2026 or early 2027, but the company acknowledges it doesn’t run in real time yet and that the compute costs required to scale to millions of concurrent players remain unsolved.

• This is an ambitious bet, but it’s so far more of an interesting demo that showcases AI R&D to shareholders than something developers or players have said they really want. Theoretically, this allows a small team to deliver a “photorealistic multiplayer experience” by bypassing AAA budgets and heavy hardware requirements; however, someone must still shoulder the underlying costs. Also, much like the recent DLSS 5 concerns, the generated imagery risks clashing with established art direction and player expectations. Perhaps it will help attract older players over time as the technology matures, but it’s important to remember that visuals are usually tertiary to what makes games successful, especially for Roblox.

Microsoft Kills Xbox Copilot AI

• Xbox CEO Asha Sharma announced that Microsoft will wind down Copilot on the Xbox mobile app and halt all development of Copilot for Xbox consoles. The feature, first unveiled at GDC 2025 as an AI gaming companion that could offer contextual gameplay advice and personalized recommendations, lasted roughly 14 months from announcement to cancellation. Sharma framed the decision as retiring features that “don’t align with where we’re headed,” while emphasizing that Xbox needs to “move faster” and “address friction for both players and developers.” Sharma herself came from Microsoft’s CoreAI division, and several of her new hires are also CoreAI veterans, suggesting this is a rejection of the specific Copilot-as-chatbot approach. She pointed to Automatic Super Resolution (AutoSR), which optimizes performance in the background, as the kind of AI integration Xbox will pursue instead. We question whether that’s the right focus either.

• Consumer-facing AI assistants don’t automatically create value just because the technology exists. Copilot for Gaming was a solution looking for a problem. Gamers already have a sprawling support ecosystem of wikis, YouTube walkthroughs, Discord communities, Reddit threads, and a chatbot overlay. AI in gaming will more likely succeed where it’s invisible and functional (graphics upscaling, matchmaking, testing, dev tools) and fail where it’s branded, visible, and competing with well-established player behavior. We’ll see if other companies like Google and Razer will continue pushing in the AI game assistant/coach direction or take the cue from Xbox.

Mindtail Raises $2 Million to Build AI-Powered Mobile Games

• Istanbul-based mobile studio Mindtail closed a $2 million pre-seed round to support what Mindtail calls an “AI-native production model,” using AI tooling across the entire development pipeline to dramatically lower per-iteration costs in the hybrid casual puzzle category. Mindtail plans to triple its headcount and announce its first game soon, with the explicit thesis that a team of seven can produce the output of a 20-person studio by building AI into the production architecture from day one.

• Hybridcasual is a hit-driven category where the economics reward rapid iteration, and the more concepts a team tests cheaply, the higher its probability of finding one that scales. If AI-native production genuinely compresses iteration costs the way Mindtail claims, it changes the competitive math of the entire category. We’ve already seen this dynamic play out in hypercasual, where studios that mastered rapid prototyping dominated until the market matured. The question is whether AI-native pipelines can deliver the same speed advantage in categories where art quality and depth actually matter. We think they can.

Other News Items

• Magic: The Gathering Arena Devs Unionize, Seek GenAI Protections

• Spotify launches “Verified by Spotify” badge to confirm artists aren’t AI

• Grok’s new open canvas is a playground for AI-generated cinema

• The new Cursor SDK is a massive threat to custom coding tools

• Claude can now plug directly into Photoshop, Blender, and Ableton

• Lovable launches its vibe-coding app on iOS and Android

• OpenAI adds animated Pets and config imports to Codex

• Google is testing new Omni model for video generation

• SAG-AFTRA secured new AI guardrails in its four-year studio deal

Content Worth Consuming

• Navigating the AI Future (Gamecraft): “I think the thesis that I’d like to operate under for this episode is that AI is going to hollow out the middle, the big middle of the game development cost stack. And it’s going to leave the top tier of developers, publishers, creatives and platforms more valuable than before. And this may have the effect of further eroding the opportunity for middle tier games because it’s going to essentially pull the ladder up to a certain extent when it comes to distribution, funding, publishing opportunities, etc.”

• The New Designer Value Proposition: How to Reposition Your Design Services in the Age of AI Art (Design Work Life): “If you feel anxious about AI replacing you, you’re not alone. As I’ve seen AI get more advanced, I’ve felt that fear myself. The key is repositioning yourself from executor to strategic director: someone who knows when to leverage AI and when human expertise is nonnegotiable.”

• How Artificial Agency’s AI companion puts personality in the hands of the player (Gamesbeat): “The demonstrations showcased the wide variety of personalities that players can easily program into Artificial Agency’s AI agent technology, including the ability to program a robot companion selfish and wily enough to leave the player character behind in a life-or-death situation.”

• AI’s Economics Don’t Make Sense(Where’s your Ed At): “If you’ve got an IT budget that was agreed at the beginning of last year, you’re getting probably 25% fewer machines for every pound spent than you were previously. The speed at which the increases have come makes it all the more difficult to bear.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — better succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, LiveOps strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.

If your organization is doing meaningful work at the intersection of AI and gaming, we still have a couple of remaining sponsor openings. Reach out if you’d like to learn more.

Now, onto today’s issue. We’re turning the lens on Krafton, one of the most deliberate AI adopters in gaming.

Krafton's Radical AI Transformation Strategy

Written by Francois Courset, Lead Consultant at Naavik

Krafton is not tiptoeing into AI. According to Kim Chang-han, Krafton’s CEO, the plan is to “establish operating standards for working in an AI-centered way and present a best practice that the global game industry can refer to.” While most publishers tread carefully to avoid alienating the vocal anti-AI gamers, Krafton was the first major publicly traded gaming company to formally adopt an “AI-first organization” mandate in October 2025. The strategy is ambitious, combining sizable investment, organizational restructuring, R&D, and consumer-facing initiatives.

Krafton’s self-appointed role as standard-bearer makes it a natural starting point. What drove this pivot? What is it actually building? And what does the ambition — successful or not — reveal about where gaming organizations are headed?

Krafton Today: Diversifying at All Costs

Krafton’s trajectory is one we’ve followed closely, and it’s been nothing short of remarkable. Built on the runaway success of PUBG, the South Korean publisher has grown into a leading force in gaming and an increasingly active investor. With a $7.9 billion market cap and strong PUBG-driven growth, Krafton has built a solid base to expand its core business and pursue new opportunities.

Its foremost challenge is to grow out of its one-trick pony status, as the PUBG IP still generates approximately 80% of total revenue. The answer lies in diversification, which underpins most company initiatives:

• Maintaining PUBG’s cultural relevance as a franchise through collaborations and IP expansions, while broadening its reach through game mode diversification and UGC deployment

• Creating new franchises through a strong internal first-party production pipeline, while scaling up the existing portfolio (InZOI, Last Epoch, Subnautica, etc.), paired with extensive capital deployment to secure future opportunities and publishing partnerships

• Geographic diversification beyond Krafton’s home Korean market and betting on future growth regions such as India

• Diversification by sectors with investments in robotics, microdramas, ad tech, and AI

The publisher is uniquely positioned to lead current AI transformations. PUBG’s war chest gives Krafton the luxury of taking chances on emergent technologies. The company is further advantaged by its Asian roots, where consumers are significantly more receptive to AI. Finally, Krafton’s young, founder-led identity with a deeply embedded engineering culture supports the agility needed to embrace large-scale transformations.

Krafton’s AI shift is unique in that it is not an isolated initiative or a PR maneuver to temporarily lift stock prices, but rather a comprehensive lens through which the company meaningfully supports all core priorities.

A Novel Shift to an “AI-First Company”

Krafton’s AI transformation plan is pioneering in many ways. It is among the leading companies to formally adopt an “AI-first” mandate — both internally and externally. The investment is also sizable: Krafton committed 100 billion won (~$68M) in infrastructure investment and an additional 30 billion won (~$20M) earmarked annually for employee training. The transformation rests on four core components.

Company Transformation: C-level Governance and Comprehensive Deployment

Central to Krafton’s organizational shift is the elevation of Kangwook Lee — formerly Head of AI — to Chief AI Officer, giving AI a seat at the top table. This move consolidates clear governance and agency to set a company-wide AI strategy, allocate resources, and pursue long-term investments needed to achieve meaningful transformation. This appointment is paired with initiatives across all layers of the organization.

The company is rolling out efficiency-oriented frameworks and setting clear benchmarks to accelerate AI adoption across the organization. Resource allocation guidelines are being defined for each development stage with the goal of cutting costs by ~40%.

On the talent side:

• Significant commitment to AI talent: As of early 2025, Krafton reported around 80 engineers focused on deep learning and AI, with plans to add about 35 more. That compares to the company’s ~2,000 headcount, ~5% of the total workforce.

• Clear mandate across organizational layers: According to Kim Chang-han: “Starting today, KRAFTON will automate work centered on agentic AI, and employees will focus on creative activities and solving complex problems as we fully implement an AI-centered management system.”

Focus on empowerment: Krafton is committing ~$20M to employee education while deploying multiple initiatives to strengthen its AI-first culture.

On the infrastructure side, Krafton is investing in a proprietary GPU cluster designed to support frontier research initiatives. While this foundation reduces the company’s external dependencies and optimizes costs, frontier research is massively capital-intensive, which could be worrying from an ROI standpoint. Moreover, while ambitious, the scale of the investment ($68M) might fall short of the requirements to compete on a global scale with frontier AI Labs.

What sets Krafton apart is the rigor of its approach. The company is systematically embedding AI across every layer of the organization, backed by clear ownership and structure.

It’s still early, but initial results are notable: up to a 16× reduction in time per task for select art workflows and reduced reliance on external UI vendors. The real test will be whether these gains hold as AI is deployed across the full production pipeline.

In-Game Initiatives: Service Mentality Over Category-Defining Innovations

Krafton is already pushing AI into player-facing features. InZoi’s early access launch in March 2025 showcased features such as face capture, video-to-motion for custom emote creation, community simulation, and autonomous characters. Looking ahead, the roadmap is anchored around “in-game AI service enhancements,” spanning improved bots in PUBG, “Co-Playable Characters” (CPCs) designed to behave like human teammates, and AI-powered anti-cheat improvements.

Deploying AI features across a game with hundreds of millions of players is a meaningful milestone, but Krafton has been notably silent on more experimental AI applications. The company was also forced to backtrack on Project Windless, reaffirming a “traditional, craftsmanship-driven approach” following player backlash. As such, Krafton’s strategy appears deliberately cautious on the player-facing side. For its first-party content, this positions the company as a fast follower focusing on already commoditized applications — like AI NPCs used by Ubisoft, Tencent, and Epic Games — instead of chasing the first breakthrough in AI-enabled gameplay. On the other hand, the company is more aggressive through its investment and publishing initiatives.

National R&D Leadership: Incentives and Frontier Research Projects

Krafton’s expanding AI division, Krafton AI, is increasingly aligning with South Korea’s state-led technology ambitions — crystallized in the AI Basic Act, which plans ~$44B in private AI investments between 2024 and 2027. Common for Korean publicly traded companies as a means to support public–private partnership and state planning, Korea’s National Pension Service holds a ~6% minority stake in Krafton.

In April 2026, the company launched “Raon,” its own AI model brand, releasing four state-of-the-art open-source multimodal models spanning speech understanding, speech chat, text-to-speech, and vision encoding.

Krafton is also a founding partner for architecture and algorithms in SK Telecom’s government-backed Sovereign AI Foundation Model project, alongside institutions such as Seoul National University, KAIST, and chipmaker Rebellions.

Fundamentally, Krafton can benefit from the government paying for part of these initiatives. This can strengthen its core gaming business (such as deepening its edge as a gateway to the Korean gaming market), while nurturing a world-class AI lab that attracts and retains AI talent. That said, such projects extend quite far from Krafton’s purview, ultimately raising the question of focus. Is there actual business merit for Krafton, or is it a societal good but a business distraction?

Beyond Gaming Initiatives: Games as a Step to Larger Market Opportunities

Perhaps most surprising is Krafton’s use of AI to venture beyond its gaming roots, which it calls “Game for AI.” This approach harnesses Krafton’s deep gaming expertise, talent, and proprietary datasets to power expansion into new verticals, robotics in particular. It’s important to note that South Korea has the world’s highest density of industrial robots per employee.

Krafton’s move in robotics is centered around two complementary initiatives:

• Internally, Krafton has founded Ludo Robotics, a robotics company developing a “social humanoid brain” capable of powering broad physical AI applications.

• Externally, Krafton is joining forces with defense giant Hanwha Aerospace through a forthcoming joint venture. This combines Krafton’s software expertise with Hanwha’s industrial capabilities in robotics, space, and military domains. As stated by Krafton’s CEO, “We aim to grow the JV into a global defense technology company like Anduril.”

The vast troves of 3D environment data held by game companies provide an unmatched foundation for training world models that learn intuitive physics, object interaction, and cause-and-effect relationships from visual and action data. In robotics, these world models allow machines to simulate thousands of actions before transferring robust, real-world-ready policies to hardware. With the proper execution, Krafton can address market opportunities beyond gaming, though the initiative is likely to concentrate capital, talent, compute, and generally the company’s focus on other ventures.

Takeaways on Approaching AI Transformation Strategies

It’s still early, but Krafton’s approach shows what a serious AI transformation looks like in practice. Krafton’s strategy stands out for its ambition and its breadth, but it isn’t a universal template. The company is pursuing multiple initiatives at once — some capital-intensive, some highly competitive, and others with unclear near-term ROI. Parts of it will likely fail, but iteration over time is part of the process.

Just as importantly, much of Krafton’s approach is context-specific. The company operates under unique conditions as a founder-led, Korean company with large and steady cash flows, enabling moves like frontier AI research, robotics expansion, and alignment with national AI priorities. Krafton’s unique circumstances — enabling it to become an increasingly unique business — don’t transfer to everyone else.

And that’s OK. The real takeaway isn’t to copy Krafton’s strategy but to learn from and match its level of commitment to the future. That’s where most companies will fall short. We’ll return to Krafton’s results in time. For now, it sets a new pace for change — one others will have to match on their own terms.

Top News Items

Written by Devin Becker, Consultant at Naavik

Roblox releases agentic AI tools, promising the ability to “build a game with a single prompt.”

• Roblox shipped new agentic AI tools for Roblox Studio, headlined by an upgraded “Planning Mode” that turns a text prompt into a structured, multi-step game design document. The AI analyzes existing code and data models, asks clarifying questions, then executes across mesh generation, procedural model creation, and even autonomous playtesting. The AI now owns the full loop: analyzing existing code, generating meshes and models, asking important questions, and running autonomous playtesting that reads logs and self-corrects bugs. Already, 44% of top creators use AI tools, and Roblox is ensuring external systems like Claude and Cursor can plug into Studio via a built-in MCP server.

• The “single prompt game” headline is aspirational marketing; the real story is the practical shift from one-off prompts to a compressed plan-review-build-test loop, more closely mirroring how professional game development actually works. This continued democratization should catalyze further content volume growth, making quality discovery more important than ever. Despite enabling additional competition among developers, Roblox is already well-positioned here as it already manages recommendations across seemingly endless experiences.

Bring NPCs to life with AI-Powered conversations in Fortnite

• Epic Games launched “Conversations,” an experimental UEFN feature that lets Fortnite island creators build AI-powered NPCs capable of real-time, unscripted voice dialogue with players. Powered by Google’s Gemini 3.1 Flash-Lite and ElevenLabs’ voice synthesis, creators define a character’s personality using simple prompts, and the NPC handles the rest. This includes remembering player actions within a session, triggering in-game events, and even adjusting difficulty on the fly. This is the productization of the same tech behind Epic’s AI Darth Vader demo. It’s still experimental, so creators can’t publish these islands live yet, but clearly Epic wants conversational AI to be a native layer of the Fortnite ecosystem.

• One challenge is handling moderation at scale. Epic has already introduced guardrails such as banning romance bots, medical/mental health advice NPCs, and filter-bypassing personas. There’s also a significant cost factor as Epic is currently eating Gemini and ElevenLabs API bills during the experiment.

Latitude unveils Voyage, a platform for creating AI-powered RPGs

• Latitude, the studio behind the original AI Dungeon, one of the first consumer products to showcase generative AI as a gameplay mechanic, unveiled Voyage. This new platform lets players design and play AI-powered RPGs. Users describe settings, quests, mechanics, and villains; then the AI generates the world. NPCs are fully unscripted, remember player actions, and can be persuaded, manipulated, or befriended. The underlying “World Engine” took five years to develop and layers multiple AI systems for narration, gameplay management, character memory, and relationship tracking. During beta, players interacted with over 160,000 unique AI-generated characters and made nearly 3,000 gameplay choices on average.

• Moderation will be challenging, and the monetization approach of free-to-play with $15-$50/month subscriptions for unlimited actions and advanced AI features will be an important test case. If Latitude can prove players will pay meaningful subscriptions for AI-driven narrative experiences, it validates an entirely new business model. We’ll have to watch and see.

Is Brave New Wonders proof there’s a place for AI in indie games?

• Brave New Wonders, a factory-automation strategy game by City From Naught, uses AI as its central gameplay mechanic. Players command AI-powered automatons using plain-text natural language instructions, including instructions to explore, build, fight, transport materials, and automate production lines. The developer is emphatic that AI is used exclusively as a game mechanic, not in asset creation. Co-founder Shala Chen noted a frustrating irony: the game’s art looks so polished that players assume it must be AI-generated, when in fact the studio chose human artists precisely because AI output couldn’t match their quality standard.

• Brave New Wonders gives players new kinds of agency and expressiveness that were previously impossible in this genre. The natural language command system creates emergent complexity that traditional programming-based automation games (Factorio, Satisfactory) can’t replicate. Using AI to let players do things they couldn’t do before is a fundamentally different proposition from using AI to do things cheaper that humans already do well. Uses like this may come across as gimmicky at first, but it only takes 1-2 breakout successes to encourage more attempts.

Other News Items

• How Tencent is using AI to make fighting games more fluid

• Why Are Large Language Models So Terrible at Video Games?

• Character.AI’s new Books mode turns reading into roleplay

• Netflix plans to add a vertical video feed, use AI for recommendations

• How AI is powering Saudi Arabia’s video games industry

• Alibaba Moves Onto Tencent’s Turf With AI Model for 3D Video

• Panic won’t release Playdate titles that use certain forms of generative AI

• Tamatem acquires Playable Factory to expand into AI-enabled ad tech

• Firstpoint VC announces €50M fund to back AI-Driven gaming and entertainment startups

• AI Vector Vision says it can fix Nvidia’s DLSS 5 to look better with new vector tech

• Adobe embraces conversational AI editing, marking a ‘fundamental shift’ in creative work

• 43% of AI-generated code changes need debugging in production, survey finds

Content Worth Consuming

• Resurrecting a 1992 MUD with Agentic AI (Metavert Meditations)

• The Future of Agentic Game Creation Roblox Tech Talks (Tech Talks)

• One of the World’s Biggest Game Makers Reveals ‘the Uncomfortable Truth’ about AI (The Game Business)

• Why Google Cloud’s Jack Buser Thinks AI Can Save the Games Industry (GamesIndustry.biz)

• Video Game Lawyer Implores Devs to Understand Ownership and Swerve Generative AI (Game Developer)

• University of Bristol Study on Player Reactions to AI NPCs (University of Bristol)

For years, Naavik has been committed to open, rigorous thinking about the game industry. Today, we’re expanding that commitment with something new.

AI x Gaming is our dedicated coverage of artificial intelligence in the games industry. We’re launching it because, frankly, we think the conversation deserves better. Despite AI becoming one of the most consequential forces reshaping how games are made, there’s no single, reliable source tracking it with depth and consistency.

Every two weeks, you can expect an in-depth perspective on a meaningful AI topic in gaming, alongside a curated roundup of the key developments you need to know about from inside the industry and beyond. This is our editorial plan in plain terms:

• Balancing the short- and long-term view — understanding where AI tools stand today, not just where they might land in five years

• Business grounding — drawing actionable insights for studios and publishers navigating AI strategy and execution right now

• Continuous coverage — returning to topics as they evolve, aligning with the fast-changing nature of AI

• Data-driven approach — making sense of the sparse, often opaque information available on how AI is actually being used

Above all, we want this to be a conversation, not a broadcast. Let’s learn together. If there’s a topic you’d like us to dig into, a perspective we’ve missed, or a question you’re wrestling with, reach out — we mean it.

One housekeeping note: this newsletter is being sent to most Naavik Digest subscribers. If you’d prefer to stick with our core Digest, you can opt out of AI x Gaming at any time.

Let’s figure this out together!

Sponsor: Create Next Generation Game Experiences with AWS

AWS for Games gives you everything you need to build, run, and grow innovative gaming experiences. Leaders across the games industry rely on AWS’s more than 16 years of games expertise to launch and run some of the world’s biggest games, including Fortnite, Roblox, and League of Legends.

AWS and over 300 AWS Games Partners offer the broadest and deepest set of purpose-built services and pre-built solutions for cloud game development, content creation, game backend & infrastructure, and player acquisition, retention & monetization. These empower game developer to build healthy global player communities, gain player insights, and personalize and monetize player experiences to build lifelong fans. 750 million gamers each month, around the world, play games built on AWS.

Learn how AWS can help you build, run, and grow innovative gaming experiences.

Learn More

Navigating AI’s Execution Era in Gaming

Written by Francois Courset, Lead Consultant at Naavik

Charting the discourse around AI’s transformative role in gaming in 2026 means making sense of multiple contradictory market signals. Within a few weeks, Krafton announced a sweeping pivot to an “AI-first” organization while Take-Two quietly dissolved most of its AI division. AI now touches over 8% of all Steam titles, yet Tencent-owned Digital Extremes made the categorical promise that “nothing in our games will be AI-generated, ever.” Over $2 billion has been directed toward game AI startups to date, yet breakout AI stories all come from other industries, Emergent being the latest vibe-coding AI startup to cross the $100M ARR mark, walking in the footsteps of Cursor, Lovable, and Replit.

Three years after GPT-4’s commercial release unleashed the AI era, the gaming industry is torn between controversies around its use in creative processes, excitement about the future efficiency and gameplay innovation prospects, and the challenges that come with proper deployment and adoption. For this inaugural issue, let’s step back from the noise to map the terrain: understand the inevitable role of AI in gaming’s future, take an honest measure of where adoption stands today, and surface the execution and governance challenges the industry will need to embrace.

The inevitable role of AI in games

While AI for games is hardly new, Generative AI marks a shift in pace and nature shifting the technology from deterministic, scripted behaviors toward open-ended systems capable of producing fully realized worlds, characters, and systems on demand. Yet what makes this leap truly significant is that it comes precisely when the industry needs it most. AI offers a powerful framework to tackle gaming’s most pressing structural challenges:

• Slow industry growth: Following the post-COVID correction, gaming has settled into a below-inflation growth trajectory, forcing publishers to rationalize costs while urgently seeking new revenue levers. With live-service models now mature and past innovations like Battle Royale no longer driving market expansion, the industry is increasingly turning to Generative AI as a new frontier: one that could unlock fundamentally new gameplay propositions and reignite growth.

• Rising production costs and need for radical production efficiency: Game production, AAA in particular, faces ballooning costs for projects yielding ever-diminishing returns. Generative AI offers a framework for streamlining everything from asset creation and QA to localization, ultimately unlocking leaner, more sustainable production economics.

• Budget pressure is further compounded by competition from emerging cost-efficient markets. Emerging production hubs across Turkey, India, China, and Southeast Asia have local talent now capable of matching quality standards from mature gaming economies, leading to compelling cost arbitrage opportunities. AI-powered tools are empowering established hubs to dramatically amplify their output, narrowing the cost-benefit of moving production to more affordable markets.

To best map out AI’s potential — not to mention structure a strategy around it — it is essential to understand its different horizons, too often mixed:

• Short-term - Efficiency: Multiple AI-enabled tools are already mature enough to be deployed and provide significant production efficiency gains.

• Medium-term - Gameplay Innovations: Initiatives to leverage AI for the core gameplay experience are still at proof-of-concept stages. Yet as first consumer-facing attempts hit the market and find commercial validation, innovations and genres will be made possible through AI.

• Long-term - Industry Disruptions: AI provides a path toward significant industry shifts, by breaking down barriers to creation, radically altering the production process itself, or through the emergence of new AI-generated entertainment mediums.

Of course, AI won’t be the silver bullet single-handedly addressing all of the industry’s shortcomings, but its resonance with many of its core concerns, while providing a horizon for future disruptions, makes it impossible to ignore. AI adoption will happen despite the mob mentality pushback from gamers (85% of gamers have a negative opinion toward AI). It might not be a smooth journey for everyone (e.g., Crimson Desert or The Alters), but at some tipping point, the gains unlocked from AI likely outweigh the pushback. And any pushback isn’t likely to be permanent either; the world adjusts and moves on, as proven by previous technology adoption patterns.

Additionally, competitive pressures will eventually accelerate the need for AI adoption. Most teams not using best-in-class tech and operating less efficiently than others will lose relevance over time. Long-term survival depends on embracing the future.

The state of AI adoption in 2026

Generative AI’s penetration into the game industry over a three-year span has been unprecedented, though exact estimates vary across reports due to methodology and sample differences. According to the GDC 2026 State of the Game Industry report, 29% of game professionals are already using Generative AI tools as part of their day-to-day workflow, while 52% report their organizations have deployed AI tools somewhere across the business.

While impressive, a closer look reveals significant room for improvement:

• First, Generative AI is far less ubiquitous in gaming versus other industries. McKinsey’s 2025 State of AI survey reported that 79% of organizations used Generative AI in at least one business function.

• Second, the nature of adoption reveals that developers are overwhelmingly gravitating toward general-purpose tools (ChatGPT, Gemini, Copilot, etc.) rather than gaming-native solutions: only 5% of developers use specialized game development AI tools.

• Finally, AI penetration is already stalling, with GDC 2026 State of the Game Industry reporting a drop from 36% of game developers to 29% over the past year. This disengagement is likely tied to growing resentment toward AI with Omdia observing that close to half of game developers anticipate that AI will negatively impact the quality of games produced, up significantly from last year.

From a product standpoint, data tells a similar story. Generative AI has already achieved significant penetration, with 8% of all Steam titles and over a fifth of new releases publicly disclosing AI tool usage — figures that almost certainly understate reality, given the declarative nature of the tracking methodology and developers’ calculated silence to avoid player backlash.

Here again, some perspective is needed: While AI-disclosed games have collectively generated an estimated $660M in lifetime revenue, this is still a drop in the ocean compared to Steam’s $16B+ yearly revenues. Notably, the top performers in this cohort skew toward indie and AA titles, with the platform’s biggest franchises largely absent. Similarly, AI use for radical gameplay innovations is still pretty much at the proof of concept stage, with few games showing actual traction (e.g. Death by AI).

Unlike other industries where Generative AI has been embraced as a productivity tool, gaming’s identity as a creative-first medium creates unique frictions on the consumer side, as it reshapes the fundamentals of how the medium is being produced: 85% of players still have a negative bias toward AI-generated content.

Finally, from an organizational perspective, a clear acceleration can be observed, with a recent string of high-profile companies setting ambitious AI partnerships: Electronic Arts with Stability or Savvy with Humain, and others radically shifting toward so-called “AI-first” organizations such as Krafton or Appsflyer. It remains to be seen whether these announcements translate into meaningful organizational evolutions as companies tackle AI deployments and evolve isolated AI Labs R&D departments into company-wide AI adoption.

To further add context to these overall adoption trends, it’s worth noting that adoption is not uniform:

• Divergence by platforms: While console/PC are slow to adopt because they walk on eggshells with the mob of gamers, mobile and its A-B testing DNA is racing forward with AI experimentation much faster, with UGC as another fertile ground for AI deployment. Examples include Supercell’s dedicated AI strategy, King’s AI Labs, or Kwalee’s early Employee Program “KAI.”

• Divergence by regions: Asia (and China in particular) is moving way faster than the West, due to a deeper product-driven approach to game development, major tech conglomerates with game production arms allowing for significant R&D investment, and fewer consumer frictions around AI usage. Examples include NetEase’s “Foundational Competency” approach to AI, or Krafton’s “prioritizing AI as a central and primary means of problem-solving.”

• Divergence by use cases: There’s obviously a big difference between using Generative AI for localization versus deeper 3D assets or player-facing gameplay interactions, so the quality of AI usage is key to track. Noting here that AI use cases go beyond solely generative ones: automate processes, performing tasks, handling decision-making logic, etc.

Far from disproving the technology’s potential, looking at adoption patterns highlights the significant road ahead for meaningful AI deployment.

The roadblocks to executing on AI’s promises

Despite robust early adoption, the industry has been surprisingly slow to produce a defining AI breakthrough, contrasting with its pioneering role in previous technological waves like VR, 3D graphics, or online social connectivity. This can largely be attributed to unique challenges tied to the nature of game production:

• Games require complex production pipelines, hosting multiple disciplines under one roof, with each subdomain having different AI tooling, quality standards, and integration challenges.

• Games’ multi-year development timelines lead to natural inertia for adopting new technologies.

• While AI’s output is typically judged on correctness, game assets are evaluated aesthetically while also requiring lore-accuracy and consistency, both creating significant overheads and limitations.

• Games are fundamentally real-time and interactive, creating challenges with computing time and latency constraints to deploy live, player-facing AI tools.

• For AI usage in gameplay, the business model is still unproven, as studios may need to put the commercial burden of computing credits on the end customer.

• As a creative endeavor, game production comes with unique cultural barriers that aren’t found across other industries. This creates additional layers of friction from gamers hating on AI use, and reluctance for employees to adopt technologies altering the very process of making games.

On top of these foundational challenges, multiple ongoing trends further complicate execution in gaming:

• As mentioned before, the industry is in a low-growth phase, which makes it increasingly challenging for studios and publishers to meaningfully commit budget to R&D while their core businesses and teams are often contracting.

• Few companies have dedicated change management teams or organized technology and tooling oversight, making it challenging to deploy experiments, but also for startups developing new tools to properly access and engage with prospective clients.

• AI talent is scarce, and the tech sector’s faster pace and deeper pockets is more competitive at acquiring talent than the gaming industry. For example, Google recently acquired three startups to bolster its DeepMind world model team, including Common Sense Machine, previously a 3D asset creation tool for games.

Once again, these roadblocks aren’t as prevalent across categories and verticals. Fundamentally, AI could be approaching another phase of its hype cycle, and as such, the narrative should shift from understanding future potential to figuring out how to actually make some of the promises work today. The road ahead is one of execution, iteration, and trial and error – and future winners will be those ready to do their homework and fully embrace these challenges.

Unlocking the next phase in AI deployment

What is the next step to deepen AI adoption in gaming?

First, a lot of the change is already happening. Mobile studios are fully embracing AI tools, Roblox has also made significant strides toward AI, and Asian companies have already radically shifted their production processes. As such, it is less about unlocking the next phase at all and rather about it being unlocked more evenly.

Second, gaming needs to find its first “Aha!” moments acting as measurable proof for the technology’s potential. Steal a Brainrot, for example, was a strong signal that gamers were ready to embrace AI-native IPs in entertainment. The industry needs to find similar validations.

Third, the AI narrative within the industry must continue to evolve. Over the last six months, companies have signaled a shift in tone — from one of secrecy and expectation management to one of clear transformation strategies and measurable initiatives — through significant organizational changes and public communications about their AI plans. However, the industry still needs the courage to proactively address and reshape gamers’ perceptions of AI. Miscommunication and poor management continue to fuel unnecessary controversy.

Gaming companies will need to evolve from isolated experiments toward clearly outlined transformation strategies, and from change management initiatives to company-wide AI governance. The current transformative potential of AI lies in a multitude of small use cases across disciplines and divisions. This requires intentionality, clear measurement, and understanding of the state of the art to identify and deliver in places where AI brings the most benefits.

Small companies should be embracing it from the ground up so they can better punch above their weight and compete with larger companies, whereas larger companies are the ones who need to radically shift their thinking and initiate large-scale transformation strategies. Inertia and willingness to innovate will define market winners for the coming years. Ultimately, the era of AI should be an era for elite small teams to achieve more than ever before and innovate in new ways that haven’t been possible in years.

Centrally, execution starts with a plan, which is exactly why Naavik launched its AI consulting vertical. If you’d like to dive deeper into AI trends, want advice on what tools and companies to work with, or need a partner to support your company’s transformation, please reach out directly or through our contact form. Let us know if we can help!

Top News Items

Written by Devin Becker, Consultant at Naavik

Owlcat Games discusses its GenAI usage on The Expanse:

• Owlcat clarified its usage of GenAI in a recent interview, stating that it is not using any AI for its existing Warhammer or Pathfinder games (both based on tabletop game franchises) but is for future games. The stated uses are for prototyping, placeholder tests, quick iteration tests, creative search/inspiration, and vision coordination before starting conceptualization.

• While it emphasized that everything in the final product will “definitely be 100 percent human-made,” there’s a potential risk in simply forgetting to remove placeholders from the final product. The most recent game to fall into this trap was Crimson Desert, which resulted in a public apology and push to replace that content. Other popular games that saw similar incidents, like Clair Obscur: Expedition 33 and Anno 117: Pax Romana, show how easily this can happen.

• Using GenAI to improve game development is not inherently wrong, but risks arise from its appearance in the released product, especially on console/PC. The online mob mentality and aggressive media coverage greatly exaggerate the issue, especially since AI-generated placeholders are less obvious than typical ones, often slipping through QA.

• Developer transparency about GenAI use, while theoretically good, can backfire, as players often misidentify human-made assets as AI-generated. This is becoming a common problem, forcing developers to issue increasingly difficult-to-prove statements to counter player suspicions, escalating tensions.

• Why does this matter to AI x Gaming: The console and PC gaming sectors face friction as AI's benefits for game creation clash with core gamers' strong rejection. Developers are attempting transparency to bridge this gap, but a prolonged period of tension is expected. Whether studios like Owlcat will face a backlash is uncertain, but the hope is that gamers will eventually adopt more reasonable expectations for AI's role in development.

Take-Two Lays Off AI Lead and Team Members

• Several news outlets recently reported on the head of Take-Two Interactive‘s AI team being laid off, along with a number of other teammates. This was reported as a major deal, especially because Take-Two had been so publicly bullish on GenAI. What was overlooked by most of the reports was that the AI group (based on the lead’s LinkedIn post) was focused on other non-GenAI forms of AI, such as procedural generation and machine learning.

• No further details have emerged about the layoff and any related internal departmental changes, so further commentary is speculative. Was it purely budget-driven? Is Take-Two leadership less bullish on this form of AI usage? Or was this structure, as a centralized team, just the wrong one for Take-Two’s diverse, sprawling business? We suspect the latter, but it’s hard to say.

• Why does this matter to AI x Gaming: Large companies often struggle to manage technological change. Unlike small teams, which are easier to get everyone moving in the same direction, large companies aren’t as nimble and have to think through structure and process to a larger, more complex degree. Take-Two clearly isn’t giving up on AI at large, but these hiccups show that transformation could prove tough for other big, traditional publishers as well. Expect more leading companies to adjust their approaches to AI in the months and years to come.

InnoGames Gives AI Life Support to Sunrise Village

• Due to declining revenue and users, InnoGames was close to shutting down its mobile F2P farming cozy game, despite it still making $160-$190k a month. Now, thanks to GenAI, the team built a strong AI pipeline to keep the live-ops going with just a 2-4 person team, while the other 20+ team members moved to a new project.

• The pipeline is built around a Unity-integrated AI Stage Designer using multiple LLMs (notably the older, deprecated OpenAI GPT-4o) to generate storylines and quests, balance progression and rewards, search and tag assets, design layouts, place objects, and create basic cutscenes in six structured steps. AI also auto-labels around 10,000 visual assets in Sunrise Village, turning what used to be hours of manual browsing into seconds by enabling semantic search like “cozy fireplace” or “autumn tree.”

• In this model, the artists aren’t replaced; instead, they are the ones prompting, curating, and correcting the AI outputs — resulting in ownership that maintains quality, reduces the need for long production periods, and frees up more time for the art team. Creating this process took roughly two months of focused engineering work, and it’s now being improved and rolled out to other projects, including those that are still in growth phases.

• Why does this matter to AI x Gaming: Players’ aversion to GenAI in games is mostly around two things: replacing human artists and a drop in quality. InnoGames shows that not only can teams avoid both of those issues, but that it’s possible to keep live service games going far more efficiently. Since mobile is in a much stronger position to capitalize on these efficiency gains, more mobile companies will integrate similar processes.

  ---

Other News Items

• Capcom says it will use Generative AI to speed up production

• Verse8 raises $5M for AI-driven game creation platform

• Sony to acquire AI and computer vision company Cinemersive Labs

• AI tech designed to end video game leaks for good uses watermarks hidden “in plain sight”

• Meta’s new model, Muse Spark, catches up on benchmarks and pivots away from open source

• The Gemini app can now generate interactive simulations and models

• Indie developer refuses to talk about the game for fear that AI will share his ideas with others

Content Worth Consuming

• “More time innovating; less time writing code:” Arc Raiders publisher Nexon believes AI will liberate its workforce (Game Developer): “Nexon CEO Junghun Lee offered more details on the company’s AI strategy, which revolves around an initiative dubbed ‘Mono Lake.’ Mono Lake is billed as an ‘end-to-end intelligence’ platform capable of weighing in on every single production decision. Nexon said the technology has been trained on billions of player sessions and decades of live game data, including engagement patterns, retention and monetisation habits, and live game telemetry. According to Lee, that vast dataset is what separates Mono Lake from other AI platforms.”

• The First AI Game Creator’s Brutal Truth About Game Dev (GameMakers): “The creator of Suck Up — the first commercially successful game using AI at runtime, with 100M+ YouTube views and $0 in marketing — breaks down what AI actually does and doesn't do for game development. Ran Mo (CEO of Proxima, ex-EA/YouTube/BCG) live demos AI-assisted coding in Unity, explains why vibe-coded games are architecturally unusable at scale, and shares his controversial take: AI in actual gameplay is probably overhyped, even by the guy who built the first AI game. We cover the trunk vs. leaves framework for deploying AI in your codebase, why the 80/100% gap in games is worth $100 million, the power law that's about to crush mid-tier studios, and why Ran hired a head of marketing with 1M TikTok followers instead of a brand marketer.”

• Project Glasswing: Securing critical software for the AI era (Anthropic): “We formed Project Glasswing because of capabilities we’ve observed in a new frontier model trained by Anthropic that we believe could reshape cybersecurity. Claude Mythos Preview is a general-purpose, unreleased frontier model that reveals a stark fact: AI models have reached a level of coding capability where they can surpass all but the most skilled humans at finding and exploiting software vulnerabilities. Mythos Preview has already found thousands of high-severity vulnerabilities, including some in every major operating system and web browser. Given the rate of AI progress, it will not be long before such capabilities proliferate, potentially beyond actors who are committed to deploying them safely. The fallout—for economies, public safety, and national security—could be severe. Project Glasswing is an urgent attempt to put these capabilities to work for defensive purposes.”

• Industrial Policy for the Intelligence Age: Ideas to Keep People First (OpenAI): “In just a few years, AI has progressed from systems capable of fast, narrow tasks to models that can perform general tasks people used to need hours to do. Now, we’re beginning a transition toward superintelligence: AI systems capable of outperforming the smartest humans even when they are assisted by AI. No one knows exactly how this transition will unfold. At OpenAI, we believe we should navigate it through a democratic process that gives people real power to shape the AI future they want, and prepare for a range of possible outcomes while building the capacity to adapt. That’s what this document is for—to start a conversation about governing advanced AI in ways that keep people first.”

More About Naavik

Naavik’s team of experts has helped over 300 companies — publishers, studios, tech companies, and investors — better succeed across the video game industry. We’d love to work with you too! Here’s what we offer, spanning all platforms, genres, and regions:

• Strategy Consulting: Projects covering market research, corporate strategy, game and economy design, gamification, LiveOps strategy, AI strategy, product management, brand and performance marketing, and more.

• M&A and Investment Advisory: Expert commercial due diligence for buyers, fundraising support for sellers, and fractional CFO/CSO services.

• Fractional Talent: The one-stop shop for top-tier fractional talent covering dozens of game industry roles — analytics, design, marketing, art, QA, and more.

Check out the links above for more details. If you’d like to chat about how Naavik can serve your team, click the box below or send us a note at consulting@naavik.co.