If you’re contemplating the use of an AI chatbot as your therapist, OpenAI CEO Sam Altman presents a reason for reconsideration. In a recent appearance on *This Past Weekend* with Theo Von, Altman recognized that the AI sector has yet to address the concern of user privacy in sensitive discussions. Unlike certified professionals, AI lacks doctor-patient confidentiality, leaving your private conversations legally unprotected.
“Individuals share the most intimate aspects of their lives with ChatGPT,” Altman remarked. “Especially younger users utilize it as a therapist, a life coach, inquiring about relationship issues and possible solutions.” However, a crucial distinction exists: “At present, when you consult a therapist, lawyer, or doctor, legal privilege applies… We haven’t yet resolved that for dialogues with ChatGPT.”
Without confidentiality safeguards, anything communicated in an AI therapy session might be accessible or subpoenaed in court. The AI industry is functioning in a legal gray area as the Trump administration manages the conflict between federal and state oversight regarding AI regulations.
Although there are a few federal statutes targeting deepfakes, the manner in which user data from AI conversations can be utilized largely hinges on state legislation. This patchwork of laws introduces ambiguity, particularly concerning privacy, which may impede wider user adoption. AI models are heavily dependent on online data for training and, in certain instances, are now being compelled to provide user chat data in legal cases.
Specifically regarding ChatGPT, OpenAI is mandated to keep records of all user conversations — including those that users have erased — due to an ongoing legal dispute with *The New York Times*. The company is contesting the court’s decision and aiming to have it reversed.
“No one had to contemplate that even a year ago,” Altman remarked, characterizing the situation as “very messed up.”
*Disclosure: Ziff Davis, Mashable’s parent company, initiated a lawsuit against OpenAI in April, claiming it violated Ziff Davis copyrights in training and operating its AI systems.*