85.8% of Japanese Game Developers Now Use Generative AI, New Survey Finds

The Numbers Behind the Headline

The headline figure is striking, but the breakdown tells an even more interesting story.

• 63% of developers say they use generative AI on a daily basis
• 22.8% use it occasionally
• Only a small minority appear to have avoided it entirely

Daily use at 63% is the number worth sitting with. This isn’t experimentation or occasional curiosity — it’s embedded in the day-to-day work of the majority of developers surveyed.

CESA’s membership includes Japan’s biggest names: Capcom, Square Enix, SEGA, Konami, Koei Tecmo, and Sony. These aren’t indie studios testing new tools on the margins. These are the studios behind some of the most commercially and critically significant games in the world.

What Are Developers Actually Using It For?

Here’s where the picture gets murkier — and more honest.

When asked about specific applications, most developers said they avoid “direct use” of AI-generated assets. The more common framing is that generative AI is used to make production pipelines more efficient, rather than to replace creative output directly.

That distinction matters, but it also raises questions. Using AI to accelerate a pipeline that produces creative work still means AI is part of the creative process — just one step removed. Whether that’s a meaningful difference is exactly the kind of debate the industry is actively having.

The Tension That Won’t Go Away

Generative AI adoption in creative industries rarely happens without friction, and game development is no exception.

The core concern is well-established: most generative AI models are trained on existing works, often without the explicit permission of the original creators. Critics argue that using these tools at any stage of production — even for efficiency — implicates studios in that broader ethical problem and gradually erodes artistic integrity.

Studios, for their part, rarely disclose in detail what their AI models are actually doing. Some don’t disclose AI use at all. The CESA survey data gives us adoption rates, but not a clear window into what “pipeline efficiency” actually looks like in practice at any given studio.

That opacity is itself part of the story.

Why This Data Matters Beyond Japan

Japan’s gaming industry is a useful bellwether precisely because it’s both creatively distinctive and commercially massive. If adoption is this high here — in an industry known for craft, long development cycles, and strong creative traditions — it signals something about where the broader games industry is heading.

The 2025-to-2026 jump from 51% to 85.8% also suggests that adoption curves in professional creative industries can move faster than most people expect once a tool clears the threshold of “good enough to be useful daily.”

For anyone tracking AI tool adoption across industries, this is a data point worth bookmarking.

What to Watch Next

The CESA survey will presumably run again in 2027. At 85.8%, there isn’t much room left to grow in terms of raw adoption numbers — so the more interesting question shifts to depth of use.

Will the share of daily users increase? Will studios become more transparent about specific applications? Will the industry develop clearer norms around disclosure?

The adoption phase appears largely complete in Japan’s game development sector. The next phase — figuring out what responsible, sustainable use actually looks like — is the harder problem, and it’s just getting started.