Generative AI is becoming more routine in game development, but it is far from universally welcomed. A new State of the Game Industry survey, based on anonymous responses from more than 2,600 developers, puts the split in unusually clear numbers: 40% of respondents say they use generative AI as part of their work, while 55% believe the technology has a negative effect on games.
The survey’s broader picture is more sobering still. Developers reported a high level of job disruption, strain associated with work and long hours, and a growing unemployment figure among respondents. Interest in Nintendo’s Switch 2 is substantial, while work on the next PlayStation and Xbox platforms remains comparatively limited for now. Union membership increased even as overall US support for unionization dipped slightly.
These results are survey responses, not a census of every studio or worker. They nonetheless offer a useful snapshot of the pressures and decisions facing people who make games: which tools are entering daily production, where platform resources are being directed, and how insecure the employment backdrop remains.
AI is in the workflow, but not beyond debate
Forty percent of respondents said they use generative AI tools in their jobs, a four-point increase over the previous survey. Of the people using those tools, 57% said their use is daily. That combination matters: the data suggests AI is not confined to occasional experimentation for a meaningful portion of adopters, but has become part of regular professional routines.
Generative AI refers here to tools that can produce or transform content from prompts or other input. The survey responses describe possible uses including repetitive tasks, research and prototyping. A prototype is an early, often incomplete version of a game, system or asset built to test an idea before a team commits to full production. In that setting, a tool that speeds up rough iteration may be evaluated very differently from one used for final creative material.
Usage is not evenly distributed. Seventy percent of respondents at mobile-only studios reported using generative AI, as did 58% of those working on free-to-play games. Free-to-play generally describes games that can be accessed without an upfront purchase, with their business model handled in other ways. The survey does not establish why adoption is higher in these groups, but it does show that AI’s role can vary sharply depending on the kind of game a team makes.
Company size also appears to affect the workplace climate surrounding the technology. Twenty-seven percent of AAA workers said they feel encouraged or pressured to use generative AI, compared with 9% of workers at indie studios. AAA is an industry label generally used for large-scale productions and larger organizations; indie commonly refers to independent studios. The wording is important: “encouraged or pressured” does not identify whether a specific tool is mandatory, nor does it prove a particular company policy. It does show that respondents at larger-scale operations were far more likely to perceive organizational momentum behind adoption.
That momentum has not erased skepticism. The 55% of respondents who see a negative impact is effectively unchanged from the previous survey’s 54%. Written responses conveyed a more complex position than a simple pro- or anti-AI divide. Some developers see potential value for limited tasks; others flagged environmental, psychological and societal concerns. Some also raised the possibility that these tools could reduce barriers for marginalized or excluded people facing constraints of cost, geography or access.
The practical takeaway is not that the industry has reached consensus. It plainly has not. Rather, game teams are likely to keep confronting questions about where AI is appropriate, who decides its use, whether it is voluntary, and how its effects are assessed across a production. The gap between daily usage among adopters and persistent negative sentiment shows that adoption alone should not be mistaken for approval.
Switch 2 interest stands apart from early next-gen console work
Among the platforms covered, Switch 2 development interest rose by almost 10 percentage points year over year, reaching 47% of respondents. That is a sizable share in the survey and places Nintendo’s newer hardware prominently in developers’ current plans and interests.
By contrast, 10% of respondents said they are currently developing for PS6, while 8% said they are working on Xbox’s Project Helix. Those figures do not say that the future systems lack projects, only that work on them has not yet spread broadly across the survey group. That is consistent with the survey’s own framing that relatively few developers are on those next-generation projects at this point.
For players, the distinction is useful when interpreting rumors and expectations. A platform can be important to a publisher’s long-range plans without being the immediate focus of most active development teams. The survey captures work reported now, rather than serving as a release schedule or a forecast of what will be announced next.
Extended-reality development appears to be an even narrower lane in these responses. Just 3% reported developing virtual reality, augmented reality or mixed-reality games. Virtual reality places a player in a computer-generated environment, while augmented reality layers digital elements over the real world; mixed reality is a related term for experiences that combine digital and physical spaces. Grouping these categories together at 3% underscores their smaller reported development footprint in this survey.
Platform planning is always only one part of the industry picture. Studios still need funding, staffing and stable production conditions to turn a target platform into a finished game. That point is especially relevant beside the survey’s employment data and its evidence that developers are navigating a difficult labor market. The industry is still producing and planning new work—including confirmed projects now moving ahead at established studios—but the workforce conditions described here complicate any easy reading of development interest as simple expansion.
Union membership rises while support slips slightly
US support for game-industry unionization was 80% in this year’s responses, down from 82% in the prior report. A two-point movement is modest, and the survey does not supply a reason for it. More notable is that the share of respondents who reported being part of a union increased from 12% to 15%.
Those two measures track different things. Support for unionization measures an attitude; union membership measures reported participation. It is possible for broad approval to change little—or decline slightly—while the number of people who are actually represented or organized rises. The results therefore should not be read as conflicting. They instead show a small dip in stated support alongside increased reported membership.
In the context of layoffs, unemployment and long hours, workplace representation will remain a consequential subject for developers. The survey does not claim that union membership solves those problems, and it cannot show cause and effect. What it does show is that participation has moved upward among its respondents during a period of pronounced employment anxiety.
Layoffs and mental-health strain dominate the human picture
The strongest warning signs in the results concern workers’ wellbeing and job security. Sixty-five percent of respondents said their mental health has suffered because of their job and long hours. This is a self-reported measure rather than a clinical assessment, but the scale is unmistakable: nearly two in three respondents connected their work conditions to harm to their mental health.
A third of respondents said they had been laid off in the last two years. Within that group, two-thirds said the layoff happened in the past year. The survey also shows how difficult re-entry can be: 54% of laid-off respondents said they had not yet found another job.
That lack of work cannot be casually characterized as a lack of effort based on these responses. Sixty-two percent of laid-off respondents said they had applied to 20 or more jobs, and 27% said they had submitted more than 100 applications. Applying for a role and being hired are very different steps, particularly when many qualified people are pursuing a smaller number of openings at the same time. The survey records the scale of that search, even if it cannot explain every individual outcome.
Across all respondents, 21% were unemployed, up from 13% in the prior report. Since the sample is made up of people who chose to answer anonymously, it should not be treated as a precise unemployment rate for the entire game workforce. Still, the eight-point change within the survey is a significant indicator of worsening conditions among its participants.
It also changes how the AI data should be understood. Tool adoption is taking place in an environment where workers are already reporting pressure, layoffs and difficulty finding another position. Developers may reasonably assess new production tools through questions of speed and access, but also through questions of control, workload, security and the human cost of changing expectations. The survey’s polarized AI result makes considerably more sense when viewed against that employment backdrop.
What this snapshot can and cannot tell us
The survey is valuable because it gathers direct, anonymous accounts from a large group of game developers, including data that would be hard to see from company announcements alone. It highlights clear patterns: rising AI use, strong Switch 2 interest, limited reported work on PS6 and Project Helix, growing union membership, and serious labor-market stress.
Its limits matter too. Anonymous respondents are not necessarily representative of every region, discipline, studio size or employment arrangement. Percentages describe the people surveyed and, in some cases, subsets such as AI adopters or people who have been laid off. They should not be stretched into claims about every developer or every company.
Even with those qualifications, the survey leaves little room for complacency. The technology story is one of adoption without settled acceptance. The platform story is one of active interest concentrated more heavily around Switch 2 than early work on the next PlayStation and Xbox systems. And the labor story is the most urgent one: a substantial share of respondents report disrupted careers, difficult job searches and work-related mental-health harm.
The next Game Developers Conference: Festival of Gaming is scheduled for San Francisco from March 1 through March 4, 2027. By then, the most meaningful changes will not simply be which new tools are more common or which hardware attracts more projects. They will be whether the people making those games see safer, more sustainable and more secure working conditions.
