According to Kotaku, Level-5 CEO Akihiro Hino posted a lengthy response on December 26 to address the online “commotion” caused by a past interview going viral. In that interview, he had suggested generative AI was writing 80 percent or more of the studio’s code. Hino now denies that claim, stating it was specifically about an unreleased title themed around AI where programmers are “deliberately” having AI handle the coding as an experiment. He argues that if a studio truly could make games with 80-90% AI-generated code, it would be “incredibly impressive.” Hino also defended AI’s role in the industry, stating it could shift AAA game development from cycles of 5-10 years down to just two years.
The Damage Control Dance
So, here’s the thing. This is a classic case of a tech executive’s forward-looking, aspirational comments getting flattened into a scary, present-tense headline. Hino isn’t exactly walking it all back—he’s reframing. The 80% figure wasn’t about Ni no Kuni or the next Layton game right now. It was a hypothetical example from a programmer on a specific, experimental project. But you can see why it went viral, right? In an industry wracked with layoffs and fears of automation, a major studio head seemingly boasting about AI doing the bulk of the coding is a five-alarm fire.
His clarification makes more sense, but it’s also a bit of a PR tightrope walk. He has to calm his own developers and the fanbase worried about soulless AI games, while still championing the technology he clearly believes in. Saying “we haven’t reached that level yet” is the key phrase. It’s an admission that the tech isn’t there for reliable, full-scale production, but it’s also a quiet promise that they’re working on it.
The Knife Analogy and Its Dull Edge
Hino pushes back against the demonization of AI, using the classic “a knife can cook or be a weapon” analogy. It’s not wrong, but in the creative fields, it feels a bit simplistic. The core anxiety isn’t that AI is a weapon, but that it’s a photocopier with a randomizer. The plagiarism fear he mentions is huge, and it’s not just about direct copying. It’s about models trained on a century of human art and code without permission, consent, or compensation, then used to potentially replace those very humans.
His heart might be in the right place—who doesn’t want “dreamlike games that surpass today’s AAA titles”? But the argument that calling AI “evil” hinders technological advancement is a common tech-bro deflection. It frames ethical and practical concerns as mere Luddite panic. The real conversation isn’t about stopping the tech, but about how it’s built, what data it eats, who profits, and who gets left behind. Dismissing that as “demonization” shuts down the debate we actually need to have.
The Two-Year AAA Pipe Dream
Let’s talk about that tantalizing promise: AAA games every two years instead of every 5-10. On one level, it’s a gamer’s fantasy. But think about what it implies. Does “saving time” with AI just mean asking the same-sized team to do even more, faster? Or does it mean smaller teams, period? And what gets lost in that acceleration? Game development’s long cycles aren’t just about writing lines of code or painting textures. They’re about iteration, playtesting, creative dead-ends, and serendipitous discoveries. Can an AI truly accelerate that process, or will it just help churn out the predictable parts faster, potentially making games feel more homogenized?
I think Hino is probably right that many studios are using AI tools quietly. Concept art brainstorming, boilerplate code, localized dialogue tweaks—these are low-risk, high-reward applications. The full statement on X and the original Automaton Media interview show a leader trying to straddle two worlds: the exciting future of tool-assisted creation and the messy, fearful present of an industry in flux. The takeaway isn’t that Level-5 is run by robots. It’s that every major studio is nervously experimenting, and the bosses are desperately hoping this particular knife is used only for cooking.
