Studio Ghibli and Other Japanese Publishers Want OpenAI to Stop Training on Their Work

Why It Matters

Compliance could force OpenAI to revamp its data‑crawling practices, impacting the pace of its generative AI development, while resistance may trigger litigation that sets precedent for AI copyright enforcement worldwide.

Summary

A Japanese trade organization, CODA, representing publishers including Studio Ghibli, sent a letter to OpenAI demanding that the company cease training its AI models on members' copyrighted works without permission. The appeal follows the popularity of OpenAI’s image generator and Sora video generator in creating Ghibli‑style content, echoing similar complaints from Nintendo and other rights holders. OpenAI now faces potential lawsuits as U.S. copyright law on AI training remains ambiguous, while Japanese law explicitly requires prior authorization for such use. The demand underscores a broader global pushback against AI firms’ “ask forgiveness, not permission” approach to copyrighted material.