Alec Radford, a former OpenAI researcher who played a crucial role in developing its AI models, has been subpoenaed in a copyright lawsuit filed against the company. The subpoena, issued on February 25, was disclosed in a court filing in the U.S.
District Court for the Northern District of California. Radford, who left OpenAI in late 2023 to focus on independent research, was the lead author of a key research paper on generative pre-trained transformers (GPTs), which form the foundation of OpenAI’s AI models, including ChatGPT. During his time at OpenAI, Radford contributed to the development of GPT models, as well as Whisper, an AI speech recognition system, and DALL-E, an AI-powered image-generation tool.
The lawsuit, titled “re OpenAI ChatGPT Litigation,” was initiated by several book authors, including Paul Tremblay, Sarah Silverman, and Michael Chabon. They allege that OpenAI used their copyrighted works without permission to train its AI models and that ChatGPT generates excerpts of their content without proper attribution.
The court previously dismissed two claims but allowed the direct copyright infringement claim to proceed. OpenAI maintains that its use of copyrighted materials falls under the fair use doctrine.
Radford is not the only former OpenAI employee facing legal scrutiny. The plaintiffs’ lawyers are also seeking depositions from Dario Amodei and Benjamin Mann, both former OpenAI researchers who left to establish Anthropic, an AI-focused startup.
However, Amodei and Mann are contesting the subpoenas, arguing they are overly burdensome. The lawsuit is one of several legal challenges OpenAI faces regarding the use of copyrighted material in AI training, raising broader questions about AI’s legal and ethical boundaries in content generation.
