Sam Altman Says ChatGPT Conversations Are Not Private Like Therapy Sessions

OpenAI CEO Sam Altman has warned users that what happens in chats with ChatGPT isn’t legally privileged in the same way that therapy, medical, or legal consultations are. The warning comes amid increasing numbers of people, particularly young users, seeking out the AI for personal advice.

In a recent Theo Von’s podcast, This Past Weekend w/ Theo Von, Altman stated that ChatGPT users frequently discuss very sensitive topics with the software, almost like a life coach or therapist. But unlike actual therapists, there are no legal safeguards surrounding such conversations.

If you discuss relationship problems or issues of mental health with ChatGPT, and there is a lawsuit, we may have to generate that,” Altman said, referencing a loophole in the way the law deals with AI interactions.

He also said chat history, despite being deleted by users, might still be accessed under certain security or legal situations. A Stanford University study has also warned of the use of AI chatbots for mental health services.

The researchers discovered that large language models tended to respond inappropriately to serious states, sometimes perpetuating dangerous stereotypes or neglecting signs of crisis. The research found that AI programs are not yet capable of playing the role of a counselor since treatment varied inconsistently from one mental condition to another and was below par in responding.

As AI becomes more dependent, particularly for emotional guidance, Altman feels that new legislation will be essential to safeguard the privacy of users. In the meantime, experts advise viewing AI as an auxiliary tool rather than one to be substituted for by professional services.