‘Talking to an AI should be like talking to a…’, says Sam Altman as OpenAI pushes back against NYT’s ‘demands’

OpenAI

OpenAI is pushing back against what it calls an “overreach” by The New York Times and other plaintiffs in an ongoing lawsuit, which demands that the company indefinitely retain all consumer data from ChatGPT and API users. The legal order, OpenAI argues, conflicts with its long-standing privacy policies and could weaken user protections across the board.

In a statement published Tuesday, OpenAI COO Brad Lightcap confirmed that the company is appealing a court order that requires it to retain user-generated content—including chats and API interactions—that would otherwise be deleted automatically within 30 days. “We give you tools to control your data… This fundamentally conflicts with the privacy commitments we have made to our users,” Lightcap wrote.

Furthermore, OpenAI CEO Sam Altman took to social media and share his thoughts about the case. “We have been thinking recently about the need for something like "AI privilege"; this really accelerates the need to have the conversation. imo [sic.] talking to an AI should be like talking to a lawyer or a doctor,” Altman wrote on X, before adding, “I hope society will figure this out soon.”

What is The New York Times seeking?

The legal demand comes as part of The New York Times’s lawsuit against OpenAI and Microsoft, accusing the companies of using copyrighted material to train AI models. As part of the discovery process, the plaintiffs are now seeking to preserve all user outputs—past and future—on the speculative basis that they may uncover evidence relevant to their case.

OpenAI has clarified that the ruling affects most consumer users of ChatGPT (Free, Plus, Pro, and Team) and API customers not using “Zero Data Retention” endpoints. Enterprise and education customers are explicitly excluded.

“We strongly believe this is an overreach. It risks your privacy without actually helping resolve the lawsuit,” the company said. OpenAI has filed motions to reconsider the preservation order and has appealed to a higher court. In a May 27 court hearing, the judge clarified that the ruling does not apply to ChatGPT Enterprise.

While the data will be retained under a legal hold in a secure, access-restricted system, it is not shared with the plaintiffs unless compelled further. OpenAI emphasised that only a small internal team can access this data, strictly for legal compliance.

The order, for now, remains in place with no end date specified. OpenAI says it will return to its regular data deletion policy if its appeal is successful.

The company reiterated that this court order does not affect model training policies, nor does it override user opt-out settings for training on chats. “We’ll keep you updated,” said the company, reaffirming its commitment to privacy and transparency.