Poll finds 100% of Japanese online game developers are using AI, though mostly for 'user preference analysis' and 'user behavior prediction'

AI is controversial for… well, any number of reasons—but particularly in game development, with studios even as beloved as Larian having to carefully walk back casual experimentations with the technology, and many developers turning their noses squarely up.

I'm not exactly a fan myself, but I'll cop that there are some mundane, often-times very boring uses where its use makes some degree of sense. I'm not about to get up in-arms about the valuable labour of resizing shoulder-pads in World of Warcraft.

Worrying my ability to ignore casual, mundane uses though is a recent report from Japan’s Online Game Association (JOGA) and Kadokawa ASCII Laboratories, as shared by Famitsu (thanks, Automaton). The poll found that 100% of Japanese developers—specifically those making online games—are using generative AI in some shape or form.

There's a bit of a silver lining, though. The poll found that most of these studios were using AI for "user preference analysis" and "user behavior prediction", rather than art, music, or voice acting. And while there are plenty of sound arguments you can make against AI adoption full stop, the use-cases here seem fairly benign as far as their impact on the creative industries go.

Details

Most doesn't mean "all", however, and I'd be deeply naive if I didn't assume plenty of these companies are using AI in ways I wouldn't approve of.

Besides, that's just from a creative, industry-impacting point of view. Am I a fan of gluts of datacenters devouring power while inflicting unsafe levels of noise pollution on residents close to where they're built? Obviously not, and I'm not entirely thrilled about industry practices that justify these things becoming standard, either.

The study found that Google Gemini had a 94% use rate, with silver and bronze medals going to Claude (84%) and GitHub Copilot (76%).

While you might assume this paints a picture of a user base that is completely ambivalent to AI, there were concerns from the players that JOGA and Kadokawa ASCII Laboratories polled, too—namely about "all games starting to look alike", so the review-tanking resistance we see on English-speaking platforms isn't happening in cultural isolation.