Game Developers Are Using AI Across the Entire Production Pipeline
Game Developers Are Using AI Across the Entire Production Pipeline
When Peter Molyneux, the veteran designer behind Fable and Black & White, spoke candidly to PC Gamer, he did not hedge. "We as the industry are pretending we're not using AI," he said, "but the reality is it's almost impossible to code now without using AI." That was not an admission of guilt. It was an operational assessment from someone who has been building games for four decades.
Molyneux went further, identifying the commercial reason most studios stay quiet. Checking Steam's AI disclosure checkbox, he noted, risks immediately alienating between 12 and 30 percent of vocal buyers. Developers are not hiding AI out of technical embarrassment. They are hiding it to protect store-page ratings from review bombing.
The statement cuts against the public posture of most technology sectors, where AI adoption is treated as a controversy to manage rather than a capability to exploit. In academia, AI-assisted work is synonymous with academic dishonesty. In the mathematics research community, AI-generated proofs have triggered serious debates about the epistemological foundations of the discipline itself. In traditional software development and DevOps, AI coding assistants draw institutional skepticism about code quality, security risk, and the long-term effects on engineering competence.
The game development community occupies a different position entirely. For them, the question is not whether AI belongs in the pipeline. The question is which part of the pipeline it has not reached yet.
From Concept to Launch
Game development is one of the most production-intensive disciplines in software. A mid-budget title requires writing, concept art, environment design, character modeling, animation, UI construction, audio production, voice acting, quality assurance, and marketing before the game ships. A major studio release layers localization, engine optimization, platform certification, and promotional campaigns on top of that. The pipeline is long, expensive, and chronically under-resourced relative to scope.
AI has found a foothold at every stage.
On the development side, coding assistants accelerate scripting, debug cycles, and engine integration work. Studios use AI tooling to build full MVP builds of scenes or game modules before committing engineering resources to a final implementation. Ubisoft's Ghostwriter system drafts NPC bark scripts — the short ambient dialogue lines that make game worlds feel inhabited — offloading an enormous volume of low-level writing work from human writers. Ubisoft also demonstrated a Neo NPC prototype at GDC and built a Teammates FPS demo using Google Gemini alongside Snowdrop engine middleware. The distance between prototype and production pipeline is shrinking fast.
On the art and content side, Square Enix confirmed AI-generated artwork in Foamstars. Level-5 CEO Akihiro Hino publicly acknowledged using generative AI to convert hand-drawn artwork into polygon models, compressing a historically labor-intensive step in the production of Professor Layton and Yo-Kai Watch titles. Concept art, UI elements, promotional assets, and marketing materials are all documented targets for generative tooling across the industry.
Voice acting and audio production are increasingly in scope as well. Steam disclosure data reflects games declaring AI-generated voice content alongside AI-generated visual assets, with studios treating both as standard production inputs.
At the strategic level, AI is entering ideation sessions, market research, and campaign planning. Krafton — the publisher behind PUBG — declared itself an "AI-first company," committed roughly $70 million to an NVIDIA GPU cluster, and partnered with SK Telecom to build proprietary foundation models. It then introduced PUBG Ally, a co-playable AI character, making its AI integration visible to players rather than buried in the production stack.
The honest accounting of AI in game development is not a list of specific tools. It is a description of the modern production pipeline itself. If a studio is using AI in one area, the data suggests they are using it in several others. The question is no longer presence. It is depth.
"We as the industry are pretending we're not using AI, but the reality is it's almost impossible to code now without using AI." — Peter Molyneux
The Studios That Said So Out Loud
By 2026, several studios had moved from quiet adoption to public disclosure.
Sandfall Interactive's disclosure to El País — revealing generative AI had been used for temporary placeholder textures during early development — cost Clair Obscur: Expedition 33 its Indie Game Awards honors, highlighting the narrow line between internal experimentation and external backlash. Pocketpair, the developer behind Palworld, addressed store transparency requirements and player sentiment publicly. Tencent Games showcased its ACE anti-cheat engine and natural-language NPC technology across Arena Breakout Infinite and Honor of Kings. Capcom is actively transforming its proprietary RE Engine into what industry analysts describe as an AI-generation game engine — integrating AI across the full production pipeline rather than treating it as a supplemental tool.
The disclosures are now institutional as well as individual. Valve's mandatory Steam AI disclosure policy prompted more than 1,000 game titles to formally declare generative AI use, ranging from indie releases up through prominent multiplayer titles like The Finals and Tribes 3: Rivals. Declarations cover assets, voice acting, UI elements, and concept art. The scope of the list reflects what developers will say officially when required to say something. The actual scope of adoption is almost certainly broader.
The numbers confirm it. A September 2026 survey by the Computer Entertainment Supplier's Association found that 85.8 percent of Japanese game developers — a group that includes Capcom, Sega, Konami, and Level-5 — use generative AI in their work. Global surveys from GDC and SlashData place adoption rates between 66 and 76 percent across professional game developers worldwide. Google Cloud research from the same period puts the figure at 90 percent.
At 90 percent adoption, this is not a trend. It is the baseline. The outlier is the developer who is not using it.
The Pragmatists
There is a structural reason why game developers landed here while other professional communities are still arguing.
Game development has never been precious about its tools. Procedural generation, licensed middleware, outsourced asset creation, performance capture studios, third-party QA farms — the industry has a long history of incorporating whatever accelerates production without tying professional identity to the choice of method. Players do not care how the NPC dialogue was written. They care whether the world feels alive.
That orientation is reinforced by the economics. Game budgets are large and development cycles are long. The gap between what a studio wants to build and what it can staff and fund to build is a permanent feature of the industry. AI closes that gap. For a solo developer or a small studio, AI tooling is not a philosophical stance. It is the difference between shipping and not shipping.
The emotional resistance that AI carries in other disciplines does exist in game development too. Player communities have pushed back on AI-generated art in specific titles. The Clair Obscur disclosure generated discussion. Some developers remain cautious about the optics of disclosure. But the operational calculus keeps running in the same direction. The tools work. The pipeline uses them.
What the game development community represents, in a technology landscape full of ideological friction around AI adoption, is a professional culture that evaluated the tools on their production merits and acted accordingly. The emotional debate did not disappear. It just did not win.
Aaron Rose is the founder and publisher of Tech Reader Magazine, covering system architecture, cloud platforms, and AI policy.