OpenAI’s Sam Altman has warned that AI tools like ChatGPT are being used as pseudo-therapists, but without the legal confidentiality protections afforded to human professionals. During a recent podcast appearance, Altman stressed that sensitive user conversations could be vulnerable in legal proceedings.

“Right now, if someone tells ChatGPT about their mental health or relationship struggles, that’s not protected,” Altman explained. “There’s no legal privilege, unlike talking to a doctor or a lawyer.”

He added that this issue is a growing concern as young users in particular turn to AI for guidance. In some jurisdictions, those chats could be requested as evidence during lawsuits or investigations.

Altman’s comments come amid OpenAI’s appeal of a court order in its lawsuit with The New York Times. The order would require it to store user chats — a move the company has called excessive and potentially harmful to user privacy.