Square Enix dropped a bombshell this week that perfectly captures everything people fear about artificial intelligence in the gaming industry. The publisher behind Final Fantasy and Kingdom Hearts announced plans to automate 70 percent of its quality assurance and debugging work using generative AI by the end of 2027. The timing couldn’t be more tone deaf, as this revelation came on the exact same day the company confirmed it was laying off over 100 employees across its US and UK offices. The message is crystal clear: we’re replacing human workers with algorithms, and we’re not even pretending otherwise anymore.
The University of Tokyo Partnership
Square Enix announced the initiative as part of a progress report on its medium-term business plan called Square Enix Reboots and Awakens. The company is partnering with the Matsuo-Iwasawa Laboratory at the University of Tokyo, a research department focused on emerging technologies that aims to create a Silicon Valley equivalent ecosystem in Japan. The joint research team includes over ten members combining university researchers with Square Enix engineers.
The project’s official name is Joint Development of Game QA Automation Technology Using Generative AI. According to the report, Square Enix wants to use automation technology to improve the efficiency of QA operations and establish a competitive advantage in game development. The 70 percent automation target by end of 2027 is aggressive considering that’s only two years away and the technology is still evolving rapidly.
What QA Actually Does
Quality assurance testers are the unsung heroes of game development. They play unfinished games for hours looking for bugs, glitches, broken mechanics, and anything that might ruin the player experience. They document every issue with detailed steps to reproduce the problem, take screenshots, and work with developers to verify fixes. It’s repetitive work that requires patience and attention to detail, but it’s absolutely critical to shipping functional games.
The job can be tedious, which is partly why some industry voices argue AI is suited for it. Automated testing tools already exist and have been used in game development for years. Major engines like Unreal include built-in automation features. But there’s a massive difference between using automated scripts to run regression tests and claiming generative AI can replace 70 percent of human QA workers. One is a tool that helps humans work faster. The other is a replacement strategy.
The Math Everyone’s Doing
If Square Enix wants AI handling 70 percent of QA work by 2027, simple logic says they won’t need 70 percent of their current QA staff. The company hasn’t explicitly stated they plan to fire QA workers, but announcing mass layoffs on the same day as the AI automation plan isn’t exactly subtle. Reports indicate up to 137 jobs are at risk in the London office alone, with cuts affecting IT, marketing, publishing, sales, quality assurance, and business planning teams.
Square Enix expects these restructuring efforts to save over 3 billion yen annually, which translates to roughly 19.6 million dollars. When companies talk about becoming lean and agile while simultaneously investing in automation technology, they’re telegraphing their intentions. The writing isn’t just on the wall, it’s in flashing neon letters.
The Industry Trend
Square Enix isn’t alone in pushing AI into quality assurance roles. An AI data firm survey from earlier in 2025 found that 31 percent of game developers confirmed they were already using AI in their workflows, with 30 percent believing AI would play an extremely important role in QA testing specifically. Electronic Arts recently announced a partnership with Stability AI to co-develop generative AI models and tools, though EA was careful to stress that human creators remain the driving force.
The difference in messaging is instructive. EA says AI can draft, generate, and analyze, but it can’t imagine, empathize, or dream, positioning the technology as a thought partner rather than a replacement. Square Enix just comes right out and says they want 70 percent automation. No sugarcoating, no reassurances about preserving jobs, just a straightforward declaration that they’re automating away most of their QA department.
Can AI Actually Do This Job
The technical feasibility is the big question mark here. Current generative AI can write code, analyze data, and produce text based on patterns it learned from training data. Some argue this makes it well-suited for identifying bugs and documenting issues. The AI could theoretically play through game sections repeatedly, spot anomalies, and generate bug reports automatically. That’s the optimistic scenario where this actually works as advertised.
The pessimistic scenario involves AI missing subtle context-dependent issues that human testers would catch immediately. Good QA isn’t just about finding crashes or obvious glitches. It’s about understanding player psychology, recognizing when something feels wrong even if it technically functions, and communicating complex issues to developers in ways that lead to solutions. An AI trained on bug reports can replicate the format, but can it truly understand why a certain enemy placement ruins the flow of a level or why specific UI choices frustrate players?
Square Enix’s History With Trends
Square Enix has a pattern of jumping enthusiastically into whatever technology trend is generating investor buzz, regardless of whether it makes creative sense. The company went hard on NFTs and blockchain gaming, launching Symbogenesis in 2023 despite widespread skepticism from both players and developers. Former president Yosuke Matsuda wrote extensively about the Metaverse playing a major role in the company’s future, right before that concept quietly died.
In February 2024, Square Enix announced it was already using Azure OpenAI during game development, with the CEO stating the company would be aggressive in applying AI. They experimented with the technology when making Foamstars, the hero shooter that launched to little fanfare and ceased updates within a year. The track record suggests Square Enix leadership is susceptible to technological hype cycles without necessarily thinking through the practical implications or long-term consequences.
The Financial Context
This AI push is happening against a backdrop of declining financial performance. Square Enix reported revenues of 133.895 billion yen for the first half of fiscal year 2025, a 15 percent drop compared to the previous year. Net income fell 14.5 percent to 10.05 billion yen. When revenues are sliding and investors are getting nervous, management feels pressure to cut costs and show they’re being proactive about efficiency.
The company also took an 11.8 billion yen restructuring expense related to reviewing development titles and overseas operations. Games like Final Fantasy XVI and Final Fantasy VII Rebirth were critically acclaimed but underperformed commercially, partly due to PlayStation 5 exclusivity limiting their potential audience. Square Enix is now pivoting to multiplatform releases and consolidating development in Japan while shrinking Western operations. AI automation fits neatly into this broader strategy of doing more with less.
What Developers Are Saying
The announcement has generated mixed reactions from the development community. Some developers acknowledge that certain aspects of QA can be tedious and repetitive, making them potentially suitable for automation. Others argue that QA requires human judgment, creativity, and empathy that current AI cannot replicate. The debate often splits along lines of whether people see QA primarily as mechanical bug hunting or as a holistic evaluation of game quality.
Workers’ rights advocates view this as yet another example of corporations using AI as justification for eliminating jobs rather than actually improving products. The Independent Workers’ Union of Great Britain has been particularly vocal following recent Square Enix firings, calling them union busting disguised as restructuring. The QA department is often among the most vulnerable to layoffs because executives undervalue their contributions until something ships broken.
The Bigger Picture
Square Enix’s AI automation plan represents a test case for how far the gaming industry is willing to go with replacing human workers. If this actually succeeds and Square Enix ships higher quality games with 70 percent fewer QA staff, other publishers will inevitably follow. If it fails and games ship in worse condition with more bugs making it to customers, hopefully there will be a reckoning about the limitations of automation.
The cynic in me suspects we’ll see a third option: games ship in roughly the same condition they do now, but with dramatically reduced development costs. Players will experience similar bug frequencies, but companies will pocket the savings from eliminated salaries. QA has always been treated as a cost center rather than a value driver by bean counters who don’t understand game development, and AI gives them the perfect excuse to slash those budgets.
FAQs
What is Square Enix planning to do with AI?
Square Enix plans to automate 70 percent of quality assurance and debugging tasks in game development using generative AI by the end of 2027.
Who is Square Enix partnering with for AI development?
Square Enix is partnering with the Matsuo-Iwasawa Laboratory at the University of Tokyo, with a joint research team of over ten members combining university researchers and Square Enix engineers.
Will this AI plan result in job losses?
While Square Enix hasn’t explicitly stated QA workers will be fired, the company announced mass layoffs affecting over 100 employees on the same day as the AI automation announcement, including positions in quality assurance.
When did Square Enix announce this AI initiative?
The announcement came on November 5-6, 2025, as part of a progress report on the company’s medium-term business plan following its latest financial results.
How much money will Square Enix save from these changes?
Square Enix expects its restructuring efforts, including the shift to AI and workforce reductions, to save over 3 billion yen annually, approximately 19.6 million dollars.
What does QA do in game development?
Quality assurance testers play games to find bugs, glitches, and broken mechanics, then document issues with detailed reproduction steps to help developers fix problems before release.
Are other gaming companies doing this too?
Yes, EA recently announced a partnership with Stability AI for similar purposes, and surveys show 31 percent of developers are already using AI in their workflows, with 30 percent believing it will be crucial for QA testing.
Has Square Enix used AI in games before?
Yes, Square Enix announced in February 2024 that it was using Azure OpenAI during development, and experimented with AI technology when making Foamstars.
What is the Matsuo-Iwasawa Laboratory?
It’s a research department at the University of Tokyo focused on emerging technologies, with a stated goal of creating an ecosystem equivalent to Silicon Valley.
Conclusion
Square Enix’s plan to automate 70 percent of QA work with AI by 2027 is either a bold leap into the future of game development or a catastrophic mistake that will result in buggier games and hundreds of lost jobs. Probably both. The technology might work exactly as promised, efficiently catching bugs and documenting issues faster than human testers ever could. Or it might miss subtle problems that ruin player experiences while executives pat themselves on the back for cutting costs. Either way, the people who currently do this work are being told their jobs have an expiration date, and that’s a grim reality regardless of how well the AI performs. The gaming industry has always struggled to value the contributions of QA professionals, and this feels like the logical endpoint of that disrespect. We’re about to find out if quality assurance is actually something that can be automated away, or if companies are setting themselves up for disaster by prioritizing efficiency over human judgment. Check back in 2027 to see how spectacularly this either succeeds or fails.