Overview
- CODA, representing publishers including Studio Ghibli, Bandai Namco and Square Enix, sent an Oct. 27–28 letter asking OpenAI to cease using members’ content to train Sora 2.
- The group says many Sora 2 outputs closely mirror Japanese films, anime and games, indicating those works were used as training data.
- CODA argues replication during the machine‑learning process may itself constitute copyright infringement under Japanese law.
- The letter also requests that OpenAI respond to members’ infringement claims and rejects Sora 2’s opt‑out policy in favor of prior consent.
- OpenAI has pledged more granular rightsholder controls, and Japan’s government has pressed the company to stop replicating local IP, but OpenAI had not commented on CODA’s request as of publication.