According to Kotaku, Square Enix announced it wants generative AI to handle 70 percent of quality assurance and game debugging work by 2027. This comes as the Final Fantasy publisher is rebooting its strategy after several flops and timed exclusivity issues with games like Final Fantasy XVI and Final Fantasy VII Rebirth. The company revealed this ambitious goal in a new investor presentation, stating they aim to “improve the efficiency of QA operations and establish a competitive advantage.” The project emerged from an internal AI-themed idea contest and will involve joint research with Matsuo Laboratory at the University of Tokyo.
Sponsored content — provided for informational and promotional purposes.
The QA automation dream
Here’s the thing – Square Enix isn’t alone in thinking AI could revolutionize game testing. A recent survey found that 30 percent of developers believe AI will play an “extremely important role” in QA testing. And honestly, it makes sense on paper. Quality assurance is brutally labor-intensive – testers might spend hundreds of hours running through the same scenarios looking for edge cases. If AI could automate the repetitive stuff, that would free up human testers for more complex analysis.
The reality check
But can AI actually understand what it’s looking at? Veteran game artist Del Walker recently pointed out a bug in Marvel Rivals where a costume completely breaks a character’s model during certain animations. “What is GenAi really gonna say when something like this pops up?” he asked. That’s the million-dollar question. An AI might detect that something looks wrong, but can it understand why it’s wrong? Can it articulate that a character’s arm is clipping through their torso during a specific attack animation at 37-degree camera angles? Probably not with the nuance a human tester would bring.
The human element matters
Look, we’ve all seen what happens when game launches go wrong. Fallout 76 became infamous for updates that somehow broke more than they fixed. QA might be at the bottom of the game development prestige ladder, but it’s often the difference between a smooth launch and a viral disaster. Companies love cutting costs, but replacing human expertise with automation in complex creative work has a track record of… well, being messy. The smart approach would be using AI as a tool alongside human testers, not as a replacement. But that doesn’t sound as impressive in investor presentations, does it?
Where this is headed
I suspect we’re going to see a lot more of these announcements as companies chase AI efficiency gains. The technology will probably get better at finding obvious bugs – crashes, performance issues, straightforward graphical glitches. But the weird, creative bugs that make games so complex? Those will likely still need human eyes. The real test will be whether Square Enix’s 2027 target leads to cleaner game launches or just cheaper QA departments. Given their recent track record, they can’t afford to get this wrong.
