100% of Japanese game devs use AI for user prediction

Artificial intelligence is stirring up controversy across many sectors, but nowhere is the debate more heated than in the world of video game development. For studios that have long been beloved, such as Larian, the rise of generative AI forces a tough reckoning, prompting many developers to pull back on casual experimentation with the technology.

The reaction from the industry has been sharp. Many professionals are increasingly turning their noses up at unchecked AI usage, fearing it threatens the valuable labor that drives creative innovation. While some see AI as an unstoppable force, others view it with deep suspicion, cautioning against letting algorithms dictate core creative processes.

Yet, amidst this skepticism, a powerful trend is undeniably underway. Despite the industry jitters, the practical applications of machine learning are proving surprisingly mundane and useful in certain contexts. Some developers are finding legitimate, if unexciting, ways to integrate AI into their workflows—even handling simple tasks like resizing shoulder-pads in games such as World of Warcraft.

This juxtaposition between industry worry and real-world adoption is powerfully illustrated by recent data from Japan’s Online Game Association (JOGA) and Kadokawa ASCII Laboratories. Their industry survey revealed a startling fact: 100% of Japanese developers creating online games are currently using generative AI in some form.

Interestingly, the focus of this adoption is less on generating art or music, and more on behind-the-scenes analysis. Most Japanese studios are leveraging AI for tasks such as “user preference analysis” and “user behavior prediction.” This usage paints a picture where AI acts more like a sophisticated data analyst than a creative partner.

While there remain valid arguments against widespread AI adoption, the current use cases seem relatively benign regarding the broader impact on the creative industries. However, this optimism is tempered by larger societal concerns. The scale of AI development inevitably leads to massive infrastructure demands, particularly within the sprawling world of data centers that devour enormous amounts of power.

Questions surrounding these operations extend beyond creative output. Developers and the public are increasingly grappling with the environmental and physical costs of this technological boom—specifically the relentless noise pollution generated by high-powered servers, which poses serious concerns for nearby residents.

Even as AI tools like Google Gemini (94% use rate), Claude (84%), and GitHub Copilot (76%) become standard tools, there remains a lingering worry among gamers about homogenization. There is concern that if AI becomes ubiquitous, games may begin to look increasingly similar, fueling fears of “review-tanking resistance” as the distinct cultural isolation of game design fades into a unified digital landscape.