adminGenerative AI disclosure is a major concern for games industry workers, with a survey finding that 88.4% believe games using the technology should identify it on digital storefronts. The responses also show broad opposition to AI-generated creative work in finished games and disagreement over whether disclosure should cover development tools as well as content players see.
Image credit: Tara Winstead
Why generative AI disclosure matters on game storefronts
A digital storefront label can help buyers understand whether generative AI contributed to a game. The survey found that support for disclosure remains high even when AI is used behind the scenes rather than to make content that appears in the final game.
A major PC game storefront introduced an AI disclosure policy in January 2024. By July 2025, around 7% of games on the platform had declared generative AI use. The policy was later narrowed so that developers were required to disclose AI used to generate content, but not AI used for development efficiency.
That distinction was not widely supported among those surveyed. Asked whether disclosure should apply only to AI content consumed by players, 48.7% disagreed with the approach, 32.1% agreed, and 19.2% answered that they were unsure or thought it might depend on the circumstances.
Many respondents said they would disclose uses beyond player-facing content themselves. A total of 70.8% would report AI use for administrative tasks, such as checking code, while 76.8% would disclose AI use during concept development. For players comparing games, a clear label or a short explanation of the tool’s role may be more informative than a simple yes-or-no notice.
What kind of disclosure do workers want?
Respondents differed on how detailed storefront information should be. A simple statement that AI was used was considered sufficient by 28.4%. Only 3.1% preferred a percentage estimate of how much of the finished product was AI-generated.
The most popular option, supported by 51.9%, was a checklist of specific uses that developers and publishers could select. Examples could distinguish between AI used for quality assurance or store descriptions and generative AI used to create concept art. Among the 13.7% who gave a different suggestion, most favoured detailed disclosure of each use.
Practical takeaway: A categorized disclosure can provide more context than a broad label. It can help readers distinguish AI-assisted production tools from generated art, writing, audio, or other content in the game.
Who took part in the survey?
The survey was open for just under two weeks and received 826 responses, including 93 partially completed forms. It was limited to people working in the games industry, so its findings describe the views of respondents rather than every developer or studio.
Most respondents worked at companies primarily focused on game development (71.2%). Two-thirds said their company did not use generative AI tools (66.1%); 30.6% said their company used them to some extent.
Individual use was less common than company-level availability might suggest. In their own projects, 78.5% said they never used AI and 9.5% said they rarely used it. A further 4.7% used it somewhat regularly, 2.9% very regularly, and 4.5% daily.
Most respondents worked for smaller organizations: 64.8% were at companies with up to 49 team members, 15.5% at companies with 50 to 250 employees, and 19.7% at companies with 251 or more employees. Art was the primary work focus for 25.2%, while 18.9% primarily worked in game design. Nearly a quarter held executive or senior leadership roles (24.8%), and 79.6% described themselves as mid-level or veterans in the industry.
Management direction on AI varied. Some 45.5% said leadership had banned or discouraged its use, while 28.0% had received no specific guidance. Another 12.7% reported some encouragement, 10.2% strong encouragement, and 3.5% said they were required to use it. These results suggest that approaches differ considerably across studios.
How generative AI is used in game development
For a question about their own uses, 602 respondents (78.1%) said they never used AI in game development. Among the uses reported, research and brainstorming led at 9.3%, followed by code generation at 9.1%, documents and reports at 7.9%, and code review at 7.7%. Other reported uses included prototyping (5.8%), translation (5.3%), scheduling and everyday tasks (4.7%), and concept art (4.2%).
Fewer respondents reported using AI for asset generation (3.5%), marketing plans (3.0%), storefront or website copy (2.7%), marketing materials (2.7%), quality assurance and testing (2.6%), or editing assets (2.5%). Reported use was also low for voice generation (2.3%), in-game text (1.8%), and music or audio generation (1.3%).
These figures can help studios create more precise internal policies. For example, teams can separately define acceptable uses for research, code review, prototyping, and material that appears in a shipped game, rather than treating every AI tool as equivalent.
Attitudes toward AI-generated art, writing, and audio
Respondents expressed strong resistance to AI-generated creative content in finished games. The shares who said AI should never be used were 86.3% for audio and music, 85.7% for in-game text and narrative, and 84.2% for art and graphics. For voice acting, 82.9% said no amount of AI use was acceptable.
In each category, around one in ten respondents considered AI acceptable for placeholder content during development, but not for the final game. Voice acting had the largest share supporting placeholder use, at 11.8%. That distinction matters for production planning: temporary assets used internally are not the same as synthetic performances or other generated material included in a release.
Different views on transparency and AI
Open-ended responses reflected a wide range of opinions. Some respondents objected strongly to AI in game development, while others described it as a tool comparable to familiar creative software or argued that smaller studios could benefit from using it. Several comments also raised concerns about training data, intellectual property, and environmental effects.
These comments are not a measure of the whole industry, and strong opinions can be more likely to appear in open-ended survey answers. Still, the survey’s numerical results point to a clear preference among respondents for transparency, alongside substantial concern about AI-generated creative assets.
What the findings mean for generative AI disclosure
The survey suggests that many games industry workers want storefront information to cover more than content players directly encounter. While respondents differ on the ideal format, the largest group favours a list of specific AI uses, which could make disclosures easier to interpret and compare.
For studios, a practical approach is to track where AI is used throughout production and separate internal assistance from content that ships with the game. For buyers, detailed storefront disclosures could clarify whether AI contributed to art, writing, audio, code, testing, or administrative work. Overall, generative AI disclosure has broad support among the surveyed workers, but expectations about what should be reported remain an important part of the debate.
