The only claim allowed to proceed is that OpenAI used copyrighted works to train ChatGPT without authors' permission, which could potentially be considered an unfair practice under California's unfair competition law. The authors have until March 13 to amend their arguments and continue pursuing any of the dismissed claims. Meanwhile, the US Copyright Office is preparing to release guidance on copyright law in relation to AI, which could impact future legal claims and potentially require works to be licensed from copyright owners for use as training materials.
Key takeaways:
- A US district judge in California has dismissed most of the claims raised by authors against OpenAI, alleging that ChatGPT was trained on pirated copies of their books without permission.
- The judge, Araceli Martínez-Olguín, agreed with OpenAI that the authors' claim that every ChatGPT output is an infringing derivative work is insufficient to allege vicarious infringement.
- The only claim allowed to proceed under California's unfair competition law alleged that OpenAI used copyrighted works to train ChatGPT without authors' permission.
- Authors have been ordered to consolidate their complaints and have until March 13 to amend arguments and continue pursuing any of the dismissed claims.