“OpenAI continues to assert that it did not willfully infringe class plaintiffs’ copyrighted works. A jury is entitled to know the basis for OpenAI’s purported good faith,” Wang wrote. “What matters is that OpenAI has put its state of mind at issue, and OpenAI may not selectively use attorney-client privilege to restrict class plaintiffs’ inquiry into evidence concerning OpenAI’s purported good faith in this way.”
As part of discovery in an ongoing lawsuit, OpenAI must release internal communications related to the mass deletion of pirated books and other allegedly copyright-infringing materials.
The authors and publishers behind the lawsuit have already gained limited access to Slack messages between OpenAI employees. Some of these messages showed workers referring to the erasure of datasets named “books 1” and “books 2,” both believed to have hosted pirated literature. However, until now, the court declined to compel OpenAI to release related communications, which the company claimed were protected by attorney-client privilege.
On Wednesday, U.S. District Judge Ona Wang issued a more definitive decision, telling OpenAI that it must present any documents that could be used to infer the company’s motive for deleting the datasets. As part of discovery, Wang said, OpenAI’s attorneys may be deposed.
The Hollywood Reporter notes that, last year, a lawyer for OpenAI said that “books 1” and “books 2” were never used for training purposes and were instead deleted in 2022 “due to their non-use.” That narrative has since been challenged by the authors’ attorneys, who believe that OpenAI likely tried to destroy evidence to protect itself from expected legal claims.
OpenAI has pushed back on the plaintiffs’ attempts to secure further records on the datasets.
For example, OpenAI initially protested requests made during discovery, saying that the communications were protected by attorney-client privilege. Later, the company moved to withdraw previous statements that the datasets had been deleted for non-use; then, it promptly claimed that any evidence directly related to the erasures constitutes privileged information.
In her ruling, Wang found that these arguments less than convincing.
“OpenAI continues to assert that it did not willfully infringe class plaintiffs’ copyrighted works. A jury is entitled to know the basis for OpenAI’s purported good faith,” Wang wrote. “What matters is that OpenAI has put its state of mind at issue, and OpenAI may not selectively use attorney-client privilege to restrict class plaintiffs’ inquiry into evidence concerning OpenAI’s purported good faith in this way.”
“OpenAI has waived privilege by making a moving target of its privilege assertions,” Wang added. “OpenAI has gone back-and-forth on whether ‘non-use’ as a ‘reason’ for the deletion of Books1 and Books2 is privileged at all. OpenAI cannot state a ‘reason’ (which implies it is not privileged) and then later assert that the ‘reason’ is privileged to avoid discovery.”
The New York Daily News notes that the two book depositories were deleted in 2022, a year before OpenAI was subject to any copyright infringement-related litigation.
“At the time, OpenAI asserted that the datasets were deleted due to ‘non-use.’ These are the only training datasets that, according to OpenAI, have ever been deleted,” Wang wrote. “Then, when class plaintiffs sought discovery about the reasons for the deletion of the Books1 and Books2 datasets, OpenAI asserted attorney-client privilege. OpenAI’s position on whether the reasons for the deletion are privileged has shifted several times.”
Sources
NYC judge: OpenAI must turn over communication with lawyers about deleted databases
OpenAI denies allegations that ChatGPT is to blame for a teenager’s suicide
OpenAI Loses Key Discovery Battle as It Cedes Ground to Authors in AI Lawsuits
Join the conversation!