If you need another reason to reconsider using an AI chatbot as your therapist, take it from OpenAI CEO Sam Altman.
In a recent appearance on This Past Weekend with Theo Von, Altman admitted to the comedian that the AI industry hasn’t yet solved the issue of user privacy when it comes to sensitive conversations. Unlike a licensed professional, an AI doesn’t offer doctor-patient confidentiality, and legally, your most personal chats aren’t protected.
“People talk about the most personal shit in their lives to ChatGPT,” Altman said. “Young people especially use it as a therapist, a life coach, asking about relationship problems and what to do.”
But there’s a major difference: “Right now, if you talk to a therapist or a lawyer or a doctor about those problems, there’s legal privilege for it… We haven’t figured that out yet for when you talk to ChatGPT.”
Mashable Light Speed
Without confidentiality protections, anything said in an AI therapy session could be accessed or even subpoenaed in court. The AI industry currently operates in a legal gray area, as the Trump administration continues to navigate the clash between federal and state authority over AI regulation.
While a few federal laws targeting deepfakes exist, how user data from AI chats can be used still depends heavily on state laws. This patchwork of regulations creates uncertainty — especially around privacy — which could hinder broader user adoption. Adding to the concern, AI models already rely heavily on online data for training and, in some cases, are now being asked to produce user chat data in legal proceedings.
In the case of ChatGPT specifically, OpenAI is currently required to retain records of all user conversations — even those users have deleted — due to its ongoing legal battle with The New York Times. The company is challenging the court’s ruling and is actively seeking to have it overturned.
“No one had to think about that even a year ago,” Altman said, calling the situation “very screwed up.”
Disclosure: Ziff Davis, Mashable’s parent company, in April filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.