Japanese Publishers Demand OpenAI Stop Using Copyrighted Content to Train Sora 2 AI Model
Summary
Major Japanese publishers including Sony's Aniplex, Studio Ghibli, and Square Enix formally demand OpenAI stop using their copyrighted content to train the Sora 2 AI model, claiming the system produces outputs that closely resemble their works without permission and violates Japanese copyright law.
Key Points
- Japan's Content Overseas Distribution Association (CODA) submits written request to OpenAI demanding cessation of unauthorized use of Japanese content for training Sora 2 AI model
- Major publishers including Sony's Aniplex, Bandai Namco, Studio Ghibli, Square Enix, Kadokawa and Shueisha confirm Sora 2 outputs closely resemble copyrighted Japanese works without permission
- CODA argues that copying during machine learning constitutes copyright infringement under Japanese law and that opt-out systems do not provide legal protection after unauthorized use