100% of online game developers in Japan already use generative AI in their productions

The use of artificial intelligence in video game development remains one of the most controversial topics in the sector. It has been documented that generative AI can have catastrophic effects on sales when the public perceives it as a substitute for creative work, and in Europe, its implementation coexists with considerable rejection within the studios themselves.

Now, a new data point from Japan reshapes the conversation: a study claims that all of the country's developers focused on online games already use artificial intelligence tools. The Japan Online Game Association (JOGA) conducted a series of surveys whose results point to adoption without exception.

The JOGA report covers video games played over the internet, whether on PC, consoles, or mobile devices, as long as they are not played entirely offline. Under this criterion, 100% of the companies working on these titles use generative AI.

The breakdown of tools is revealing: Google Gemini appears in 94% of the surveyed developers, Claude in 84%, and GitHub Copilot in 76%. The sum far exceeds 100%, a sign that studios combine multiple models depending on the task instead of marrying themselves to a single provider.

The primary stated goal is not to generate art or write scripts, but to analyze user preferences and predict behaviors. In other words, AI enters through the business door—retention, monetization, segmentation—before that of content.

In general development, beyond the online segment, adoption is advancing primarily to avoid routine tasks and improve work efficiency. The study also points out that many studios keep it discreet: they are already using AI even if the player does not perceive it.

The survey also assessed public perception, and concerns were concentrated on two fronts. The first: that games violating intellectual property laws will proliferate by being trained or generated using third-party material.

The second, perhaps more corrosive in the long term: that games will begin to look increasingly alike. If all studios optimize using the same models and the same behavioral data, the homogenization of the catalog ceases to be a theoretical risk and becomes a statistical consequence.

The Japanese case serves as a mirror for iGaming and digital entertainment in general. When adoption reaches 100%, AI ceases to be a competitive advantage and becomes infrastructure: no one wins by using it, everyone loses for not doing so.

From that point on, differentiation returns to where it has always been: in editorial judgment, creative direction, and the ability to decide what not to automate.