Ghibli, Bandai Namco, Square Enix urge OpenAI halt AI; Sora opt‑out breach law
When Studio Ghibli, Bandai Namco and Square Enix realized that OpenAI’s Sora had been fed thousands of frames from their anime series and game cutscenes, they each sent a formal letter demanding the training stop. The studios note the wider AI-rights debate, but they point out that Japanese law treats the use of protected material differently from the “fair-use” defenses you hear about elsewhere. Their counsel, CODA, says the very opt-out system OpenAI offers might itself clash with domestic copyright rules - basically, you need clear permission before any content gets repurposed for machine learning. That clash between OpenAI’s opt-out approach and Japan’s statutory requirements sits at the core of the fight, and it makes you wonder if the company’s method can hold up under Japanese law.
Sam Altman said last month that OpenAI will tweak Sora’s opt-out policy for rights holders. CODA, however, argues the policy’s existence may already have broken Japanese copyright rules, noting that “under Japan’s copyright system, prior permission is generally required for the use of copyrighted material.”
Altman announced last month that OpenAI will be changing Sora's opt-out policy for IP holders, but CODA claims that the use of an opt-out policy to begin with may have violated Japanese copyright law, stating, "under Japan's copyright system, prior permission is generally required for the use of copyrighted works, and there is no system allowing one to avoid liability for infringement through subsequent objections." CODA is now requesting on behalf of its members that OpenAI "responds sincerely" to its members' copyright claims and stops using their content for machine learning without their permission, which seems to include not just Sora output, but also the use of Japanese IP as training data.
CODA’s letter makes it clear they expect OpenAI to comply. The trade group, which speaks for Studio Ghibli, Bandai Namco, Square Enix and a few others, says the way Sora 2 was trained could be copyright infringement because the model has already spitted out scenes with protected characters. Altman announced a tweak to Sora’s opt-out policy for IP owners last month, but CODA argues that even an opt-out runs afoul of Japanese law, which usually demands permission before any use of copyrighted material.
That raises a tricky question: does an opt-out fit Japan’s legal framework, and will OpenAI’s new approach satisfy the association? Without a court decision, the legal weight of the replication claim stays fuzzy. OpenAI hasn’t laid out exactly how the revised policy will work, nor how it will meet CODA’s specific worries.
As the back-and-forth continues, both sides seem ready to push the limits of copyright enforcement for generative AI, but the final outcome remains uncertain.
More in Policy & Regulation
Congress adds energy provisions to defense bill to spur advanced reactors
Why does a defense bill suddenly matter to the power grid? While the National Defense Authorization...
Spain mandates AI‑generated dashboards, reports, and slides be labeled
Spain has moved from discussion to enforcement, demanding transparency for every piece of analysis...
Living policy records bring real‑time AI ethics visibility across departments
The 2026 outlook for AI ethics and governance is shifting from paper‑based rulebooks to dynamic,...
Google Translate adds real‑time speech translation to any headphones
Google’s latest update to its translation service expands the reach of spoken‑language assistance...
After ditching AI fitness apps and a Fitbit, I return to Peloton classes
The surge of AI‑powered workout assistants has turned many home gyms into data‑driven command...