Key Notes
- Forty percent of surveyed game-industry professionals use generative AI at work, up from 36% in the previous survey.
- Negative sentiment remains broadly stable at 55%, compared with a corrected 54% previously, despite higher adoption.
- Usage reaches 70% among respondents at mobile-only studios, while research and brainstorming remain more common applications than final game assets.
Forty percent of game-industry professionals surveyed by GDC now use generative AI at work, up from 36% in the previous survey. But increased use has not brought broad approval: 55% believe the technology is harming the industry, compared with a corrected 54% previously.
The November State of the Game Industry report from GDC draws on more than 2,600 respondents across 69 countries. Its findings describe a workforce finding uses for AI while remaining concerned about jobs, creative control and the consequences of relying on generated work.
GDC characterizes sentiment as broadly stable. The one-point rise in negative views is much smaller than the shift since 2024, when 30% gave that answer, and sits within the report’s stated overall margin of error of plus or minus three percentage points.
Mobile-Only Studios Lead AI Adoption
Usage is far from uniform across the business. Seventy percent of respondents working at mobile-only studios said they use generative AI, compared with 58% of those working on free-to-play games and 32% of those making premium games.
The mobile figure describes people at studios devoted to mobile, rather than every developer who has worked on a phone game. These categories can also overlap: a mobile studio can make free-to-play games, so the percentages should not be added together.
Among respondents already using AI, 57% said they use it daily. Adoption also differed geographically. US usage moved from 36% to 38%, which GDC says is within its margin of error, while usage outside the US rose from 37% to 44%.
Company expectations vary too. Twenty-seven percent of AAA studio workers felt encouraged or pressured to use generative AI, compared with 9% at indie studios. That combined category does not distinguish a welcome recommendation from pressure an employee resents.
Coding Is Common, Finished Assets Are Not
The report’s task breakdown helps explain why using AI and opposing its wider impact can coexist. Research, brainstorming and summarizing information were more common than creating material that ends up in the finished game.
| Task | Share of AI adopters |
|---|---|
| Research and brainstorming | 73% |
| Data analysis and summarization | 62% |
| Code generation and debugging | 56% |
| Proofreading and editing | 42% |
| Slide decks and presentations | 39% |
| Prototyping | 37% |
Only 8% of AI adopters reported using it for production art or final game assets, and 7% for player-facing features or content. Those are shares of AI users, not of everyone surveyed, and the questions measure reported usage rather than the amount of generated material in a game.
Tools are also spreading into specialized work. We covered Aura Autopilot, an open-source agent designed to playtest projects in Unreal Engine, Unity and Godot. Its release illustrates a possible use; the GDC survey does not establish how well any particular tool performs.
A Majority Still Sees Harm
Alongside the 55% reporting a negative impact, 35% described AI’s effect as mixed and 8% as positive. Quality assurance and community support workers were the most negative group at 76%, followed by game design and visual or technical arts, both at 65%.
Respondents raised concerns about environmental costs, legal questions and job losses. GDC also identified worries that relying on AI could weaken professional skills. Some people who judged the overall impact negatively still saw value in using it for repetitive work or early experimentation.
We previously reported employee accounts of EA’s AI push arriving alongside smaller teams and continuing workload demands. Those accounts provide workplace context; neither they nor this survey establish that AI caused a particular round of layoffs.
The sample matters when reading the totals. US residents represented 63% of respondents, up from 54% previously, and GDC says the survey attracted greater participation from people affected by layoffs than other employment estimates suggest. The report also discloses selective use of internal AI tools during research and editing, including sentiment analysis. These results describe the surveyed professionals, rather than a census of the global development workforce.
