6 months after revealing plans to become 'aggressive in applying AI', Square Enix's president announces he's being careful in applying AI

Aerith, a peppy and optimistic character in Final Fantasy 7 Remake, looks a tad confused towards the right of the screen.
(Image credit: Square Enix)

Square Enix appears to have learnt its lesson when it comes to sparkly new technology—just with a considerable input delay. Earlier this year, Square Enix's president Takashi Kiryu spoke glowingly of generative AI, which carried strong echoes of the company's equally ill-timed love affair with NFTs.

In a newsletter from January, he stated: "I believe that generative AI has the potential not only to reshape what we create, but also to fundamentally change the processes by which we create," and that the company would be "aggressive in applying AI and other cutting-edge technologies to both our content development and our publishing functions."

As reported by michsuzu on Twitter (and later spotted, then translated by Automaton), in a June 21 call to investors, Kiryu seems to believe that, just maybe, being hyper-aggressive about applying a controversial new technology isn't the best of ideas:

"AI itself has tremendous potential. However, there are also many risks involved." Said risks are not just relegated to the court of public opinion—generative AI technology has a habit of hallucinating information. For instance, it may completely fabricate a recipe, invent a case study in a lawsuit, or create a coding solution that doesn't exist in programming. "We have introduced a flow whereby AI-related tools are used internally only after being properly examined," Kiryu added.

Looking at the wider industry, most actual developers aren't super enthusiastic about its inclusion either, outside of confining it to boring grunt work—basically, the scope of the tech is there, but limited, and often decided on by the people actually making the stuff rather than far-removed executives who think it could, like, totally add a billion immersions and drive up emotions by 500%.

Kiryu seems to have actually digested these criticisms properly, though, stating that the company would only be applying generative AI to "areas unrelated to creativity"—better late than never, as they say. He notes that, in particular, using generative AI in creative fields is "extremely delicate", and that he'd be keeping an ear to the ground.

Overall, a refreshingly even-keel answer from Kiryu—personally, I'd prefer to have avoided Square Enix's confusing PR blunder with Foamstars (and whatever the hell Symbiogenesis is), but it's nice to see the president of a giant like Square Enix step outside of the deep learning hype sphere, even if it's just for a second.

Harvey Randall
Staff Writer

Harvey's history with games started when he first begged his parents for a World of Warcraft subscription aged 12, though he's since been cursed with Final Fantasy 14-brain and a huge crush on G'raha Tia. He made his start as a freelancer, writing for websites like Techradar, The Escapist, Dicebreaker, The Gamer, Into the Spine—and of course, PC Gamer. He'll sink his teeth into anything that looks interesting, though he has a soft spot for RPGs, soulslikes, roguelikes, deckbuilders, MMOs, and weird indie titles. He also plays a shelf load of TTRPGs in his offline time. Don't ask him what his favourite system is, he has too many.