The legal landscape surrounding generative AI and copyright remains unsettled, with ongoing disputes in courts. Studio Ghibli could potentially bring two types of copyright cases against OpenAI: one concerning the use of its works for training AI models ("input" case) and another regarding the creation of images resembling its copyrighted products ("output" case). However, proving these claims could be difficult, especially if OpenAI's models were trained on fanart or if the generated images merely evoke the style rather than replicate specific works. While Japan has permissive copyright laws for AI training, Studio Ghibli could still pursue legal action in the United States.
Key takeaways:
- OpenAI introduced a feature allowing users to generate images in the style of Studio Ghibli, raising copyright concerns.
- Studio Ghibli may struggle to win a copyright lawsuit against OpenAI, as general styles are not protected by copyright law.
- Legal arguments against OpenAI could focus on whether its models were trained on Studio Ghibli's works or if the generated images resemble copyrighted products.
- While Japan has permissive copyright laws for LLM training, Studio Ghibli could still pursue legal action in the United States.