OpenAI CEO Warns ChatGPT Conversations Aren’t Legally Confidential
In a candid podcast appearance, OpenAI CEO Sam Altman has issued a critical warning to users of ChatGPT: conversations with the AI chatbot are not protected by legal confidentiality.
Speaking on comedian Theo Von’s podcast This Past Weekend, Altman acknowledged that many users—particularly young people—are treating ChatGPT like a therapist or life coach, often sharing deeply personal details. However, he emphasized that unlike licensed professionals such as lawyers, doctors, or therapists, AI platforms do not offer legal protections like doctor–patient or attorney–client privilege.
“People talk about the most personal details in their lives to ChatGPT,” Altman said. “And we haven’t figured that out yet for when you talk to ChatGPT.”
Altman went on to express concern that in the event of legal disputes, user conversations with the chatbot could potentially be subpoenaed or otherwise accessed.
“If you go talk to ChatGPT about your most sensitive stuff and then there’s like a lawsuit or whatever, we could be required to produce that—and I think that’s very screwed up,” he warned.
Call for “AI Privilege”
Altman floated the idea of developing a legal framework for “AI privilege”—a form of confidentiality that would protect user conversations with AI platforms similarly to communications with licensed professionals.
“I think we should have the same concept of privacy for your conversations with AI that we do with a therapist,” he said. “No one had to think about this even a year ago.”
Legal Pressure Intensifies Privacy Concerns
Altman’s remarks come as OpenAI faces mounting legal scrutiny. In a copyright lawsuit filed by The New York Times, a U.S. court has ordered OpenAI to preserve and segregate all ChatGPT output, even those chats users believed had been deleted.
U.S. Magistrate Judge Ona T. Wang issued the order on May 13, 2025, and District Judge Sidney Stein upheld it on June 26. The ruling requires OpenAI to retain ChatGPT conversations for users across Free, Plus, Pro, and Team accounts. Enterprise and educational accounts are currently exempt.
This development temporarily halts OpenAI’s usual practice of deleting chat data within 30 days and has prompted renewed calls for stricter privacy regulations.
Unlike encrypted messaging services such as Signal or WhatsApp, ChatGPT conversations are not end-to-end encrypted. OpenAI’s privacy policy explicitly states that user data may be shared with third parties, including government authorities, when required by law or to prevent harm.
Experts Urge Caution
Privacy experts and cybersecurity analysts are echoing Altman’s warning, advising users not to share legally or emotionally sensitive information with AI chatbots. Until new legal protections are enacted, AI interactions should be treated as unsecured digital communications.
“This is a wake-up call,” said one cybersecurity analyst. “Until we have concrete legal protections, treat AI like a public forum—not a private diary.”
OpenAI has not yet issued an official statement addressing Altman’s comments, but the debate over AI privacy is expected to escalate as regulators and industry leaders explore new frameworks to protect user data in the rapidly evolving age of artificial intelligence.
