AI in Games Is Already Here. You Already Accepted It.
The debate about whether AI should be used in game development is already over.
ReaperX
·
28 April, 2026
·
3 min read
The debate about whether AI should be used in game development is already over. It is in use. The question worth asking is what it is being used for, who benefits, and whether the games made with it are better or worse than the ones made without it.
The GDC 2026 State of the Game Industry report found that 52% of game professionals believe generative AI is harmful to the industry. That is a majority of the people making games concluding that a tool their employers are deploying is bad for their field. The 85% of gamers who hold negative attitudes toward AI in games, per Quantic Foundry research, mostly have a different concern: they do not want AI-generated content in the games they pay for.
Both groups are responding to the same underlying problem. AI in games is being deployed primarily as a cost-reduction mechanism rather than a creative one.
7000 Steam titles disclosed AI use in 2025
Ubisoft has used AI to generate NPC dialogue. Several studios have used AI to generate concept art in pre-production. According to analysis from Whimsy Games, over 7,000 Steam titles disclosed AI use in 2025, representing roughly a third of all releases. The same analysis coined the term “gameslop” for the low-effort titles assembled primarily through AI tools with minimal human curation, which flooded storefronts and damaged discoverability for everything around them. The player response to gameslop has been measurable. Review scores for AI-heavy releases skew negative, not because players run content audits but because the quality ceiling of current generative tools sits below what careful human work produces at its median, and players notice.
The Skill Issue is the assumption that AI use is binary: either you are against it and therefore a Luddite, or you are for it and therefore practical. That framing serves publishers who want to reduce headcount without the argument. The actual distinction is between AI as a tool that expands what a team can produce, and AI as a replacement for the people doing the producing. The first is defensible. The second is what 52% of developers are observing in practice.
The narrative that AI will democratise game development is true at the level of small-scope projects and misleading at the level of the games the industry is actually cutting staff to produce. A solo developer using AI to generate placeholder art while building a prototype is using the tool well. A AAA publisher cutting its concept art team and using AI generation for shipped assets is making a different decision and should be honest about it.
Negative attitudes
What players are responding to when they express negative attitudes toward AI in games is the output quality problem. AI-generated content is recognisable at the current level of tooling. It has specific failure modes: repetitive structures, tonal inconsistency, a quality ceiling that sits below human craft at its median. The 85% negative sentiment figure exists because people have encountered the output and found it wanting.
The publishers cutting creative staff and replacing them with AI generation are betting that players will not notice, or will not care enough to stop buying. The data from 2025 suggests that bet is wrong. The games that performed best critically and commercially were the ones with authored content: Baldur’s Gate 3’s continued cultural footprint, Monster Hunter Wilds’ world-building, games where human creative decisions were visible in the product.
AI in games is not going away. The Skill Issue is treating its deployment as inevitable rather than as a series of choices that specific studios are making, choices that can be assessed, criticised, and factored into which studios deserve support.