Ghibli Challenges OpenAI's AI Training

Studio Ghibli, along with other Japanese publishers, is demanding that OpenAI cease using their copyrighted content to train AI models like Sora 2. The group, represented by the Content Overseas Distribution Association (CODA), alleges that OpenAI is exploiting a legal loophole in Japan's copyright laws. This allows the incorporation of anime, games, and other cultural works into AI training datasets without explicit permission.

The demand specifically targets OpenAI's video-generation tool, Sora 2, urging an immediate halt to training involving protected content. Concerns arose earlier when ChatGPT's image-generation features mimicked Studio Ghibli's style. Japan's copyright revisions in 2018, intended to promote AI innovation, inadvertently created a loophole that benefits foreign firms like OpenAI, while restricting local creators.

CODA is seeking governmental intervention to protect national creative assets and requests that OpenAI respond sincerely to claims and inquiries from CODA member companies regarding copyright infringement related to Sora 2's outputs. This action highlights the growing tensions between creative industries and AI developers regarding copyright and the ethical implications of AI training practices.