Square Enix is currently in the middle of a reboot. Following a bunch of flops (remember Foamstars?) and some great games held back by timed exclusivity (Final Fantasy XVI, Final Fantasy VII Rebirth), the RPG publisher has been pivoting to a multiplatform approach and retooling its slate of upcoming releases. It’s also, apparently, making a big push into leveraging generative AI. The company announced a new goal of having 70 percent of quality assurance and game debugging work handled by generative AI by 2027.
“Through the use of automation technology, [we] aim to improve the efficiency of QA operations and establish a competitive advantage in game development,” reads a new presentation updating investors on Square Enix’s business strategy (via VGC). The project developed out of an AI-themed “idea contest” the company held, and will be part of a joint research initiative with Matsuo Laboratory at the University of Tokyo. Is this a real plan or just a way to impress investors with buzzwords?
Square Enix isn’t the only company trying to apply generative AI tools to QA. While some companies get lampooned for letting AI slop infest concept art or even finished loading screens that appear in the game, others have argued that the job of discovering and documenting bugs is one where LLM technology will be most useful. An AI data firm claimed earlier this year that 30 percent of respondents to a developer survey believed AI would play an “extremely important role” in QA testing.
I’m convinced that statement from Square is to just get shareholders excited, because what is GenAi really gonna say when something like this pops up? Can it even critically understand what it’s looking at to report the issue accurately? pic.twitter.com/VJ5r67Fu27
— Del (@TheCartelDel) November 6, 2025
But will it actually be any good or save studios time and money? Veteran game artist Del Walker pointed out a recent bug in Marvel Rivals by which a costume completely breaks a character’s model during certain animations. “What is GenAi really gonna say when something like this pops up?” he wrote on X. “Can it even critically understand what it’s looking at to report the issue accurately?”
Quality assurance has long been at the bottom of the hierarchy when it comes to game development prestige and pay, despite bugs and glitches being the first things to go viral during a game launch, setting the tone of how big-budget games are perceived. They can also make or break its initial reception if things are particularly bad. Fallout 76 was particularly bad when it first released and was infamous early on for important updates that found new ways to accidentally break the game.
QA, in other words, is very important, but it can also be very labor intensive. If we know one thing about companies its that they love results and hate having to pay people to produce them. It’s easy to envision a way forward where developers on the studio floor are able to experiment with new tools and see whether they work before committing to ambitious goals that replace expertise with automation. That probably doesn’t sound as good in a PowerPoint presentation, though.
