Nihon Falcom Corporation, the mid-sized Japanese studio best known for The Legend of Heroes and Trails series, has shed some light on its internal 'generative AI' strategy during a recent shareholders' meeting. While they are adamant about not using these tools for the final product delivered to players, they admit to using them for brainstorming scenarios, gathering data, and proofreading. It has helped them significantly reduce development time.

The details emerged during the Q&A session at Falcom’s 2025 shareholders' meeting held on the 18th. When asked about their internal use of generative AI, the studio clarified, "Due to legal concerns such as copyright, it is still difficult to use it directly in our products. Currently, we are limiting its use to internal purposes, such as data collection and internal meeting materials." Specifically, they noted that AI is being used to brainstorm scenario ideas and proofread text.

Falcom also added that there is a significant time-saving effect. They mentioned that some tasks which previously took two to three hours can now be completed in just 10 minutes. Based on this, they stated that they believe AI is effective for reducing development time and costs.

In summary, while Falcom cannot directly incorporate results from generative AI into their games due to legal issues like copyright, they are seeing positive effects in terms of time savings for internal operations such as company meetings.

Japan, a powerhouse in the industry, is seeing a steady rise in AI adoption. According to a report by the Computer Entertainment Supplier's Association (CESA), 51% of the 54 Japanese game companies surveyed are already using AI. Key use cases include generating visuals and character art, drafting stories and text, and assisting with programming—with 32% of firms even developing their own game engines using AI. Square Enix has set a target to automate 70% of its QA and debugging with AI by 2027, and Sega is also looking to improve efficiency by streamlining development projects through AI.