Studio Ghibli-backed group urges Sam Altman's OpenAI to stop using Japanese art in Sora 2 AI training

Studio Ghibli-styled AI-generated images on OpenAI's ChatGPT feature latterly suit a sense impression among netizens for mimicking the studio's signature tune art style. However, Studio Ghibli's co-founder, Hayao Miyazaki, a staunch critic of AI in the creative process, said such AI-generated images clash directly with the artistic philosophy.

Now, a leading Japanese trade organisation, representing major publishers including Studio Ghibli, has formally called on the Sam Altman-led OpenAI to cease using Japanese copyrighted content in the training of its artificial intelligence systems. The appeal was made last week.

The Content Overseas Distribution Association (CODA) sent a written request to OpenAI on 27 October, raising concerns about the company’s video-generation platform Sora 2 (sora.chatgpt.com), launched on 30 September.

CODA said it had confirmed that “a large portion of content produced by Sora 2 closely resembles Japanese content or images,” and believes that such resemblance indicates the use of Japanese works as training material for the model.

Potential copyright infringement under Japanese law

In its letter, CODA argued that reproducing copyrighted works during the AI learning process could amount to copyright infringement, especially when generated outputs resemble or replicate existing creative material.

The association also took issue with OpenAI’s opt-out system, which allows copyright holders to request the removal of their works from training data after the fact. CODA said this approach conflicts with Japan’s copyright framework, which generally requires prior authorisation for the use of protected works.

CODA’s letter outlined two key demands:

The organisation urged OpenAI to work towards a balance between technological innovation and creators’ rights, saying both should be protected to ensure “the healthy development of AI technology.”

The development comes weeks after OpenAI faced criticism for its handling of sensitive content on Sora. The company recently paused the generation of images depicting Martin Luther King Jr, following objections from the civil rights leader’s estate.

In a statement posted on X on 17 October, OpenAI said it had worked with The Estate of Martin Luther King, Jr Inc to ensure the late leader’s image was treated respectfully. The decision followed reports that some users had created “disrespectful depictions” using the AI platform.

“At King, Inc.’s request, OpenAI has paused generations depicting Dr King as it strengthens guardrails for historical figures,” the company said.

The letter from CODA adds to a growing international debate over how AI companies source and use creative content to train their models. Governments and rights organisations across the world are pushing for greater transparency and explicit consent from creators whose works underpin these systems.

---

---