OpenAI is fighting to keep cofounder Ilya Sutskever’s files secret
OpenAI pushes against California’s AI bill which requires a “full shutdown” feature from AI companies.
OpenAI apparently doesn’t want anyone poking around its former co-founder’s computer.
Lawyers in a high-profile copyright lawsuit between OpenAI and book authors are fighting over whose files should be subject to discovery.
In a heavily redacted letter to the judge filed on Thursday in New York federal court, lawyers for the Author’s Guild said OpenAI has objected to including six current and former employees as “custodians” in the lawsuit. A custodian refers to someone in possession of relevant evidence that would need to be turned over during the discovery process.
According to the letter, lawyers from both sides have met multiple times and agreed upon a list of 24 custodians but have reached an impasse on these six names, which include co-founder and former chief scientist Ilya Sutskever, pretraining data lead Qiming Yuan, technical staff member Jong Wook Kim, technical staff member Shantanu Jain, former research scientist Cullen O’Keefe and former science communicator Andrew Mayne. Mayne has written publicly about the importance of books as training data for LLMs.
Sutskever was instrumental in orchestrating the firing of CEO Sam Altman in November 2023 and has since left the company to start a competing AI model developer.
Lawyers say in the letter they believe he is in possession of documents relevant to the lawsuit. OpenAI, for its part, has not said publicly why it opposes including Sutskever and the five others in the discovery process. Now, lawyers for the Author’s Guild are asking the judge to step in and compel OpenAI to include these names as custodians.
All six of the exhibits attached to the letter have been fully redacted at the request of OpenAI. The company said in court documents that it requested the redactions because the exhibits include proprietary source code as well as “discussions between OpenAI employees describing detailed processes for training and testing ChatGPT models.”
The lawsuit is one of many alleging that OpenAI violated copyright when it used millions of books to train its AI models. Plaintiffs include authors such as George R. R. Martin, Jonathan Franzen, and David Baldacci.
