Studio Ghibli, Bandai Namco, Square Enix Demand OpenAI Stop Using Their Content to Train AI

Why It Matters

The dispute underscores a growing legal battle over AI training data, potentially forcing AI firms to obtain explicit licenses for copyrighted content and reshaping how generative models are developed worldwide.

Summary

The Content Overseas Distribution Association (CODA), representing Japanese IP owners such as Studio Ghibli, Bandai Namco and Square Enix, sent a formal letter to OpenAI demanding that the company stop using their copyrighted works to train its text‑to‑video model Sora 2. CODA argues that the replication of protected characters during machine‑learning constitutes copyright infringement and cites Japan’s copyright law, which generally requires prior permission. The request follows a Japanese government appeal for OpenAI to halt the generation of Japanese‑style content after Sora 2 produced a flood of images featuring the studios’ characters. OpenAI has pledged to revise Sora’s opt‑out policy, but CODA says the original use of its members’ material was unlawful and urges OpenAI to cease all training on Japanese IP.