Your conversations with ChatGPT could be used against you in court

The Washington Post revealed that conversations users have with artificial intelligence chatbots, particularly ChatGPT, could become evidence and documents used in courtrooms and criminal investigations, despite users’ sense of privacy while chatting with the bot. The newspaper, through a review of public records and local reports, identified the use or reference of AI chatbot conversations in 12 cases over the past two years, raising questions about the privacy of these exchanges and their potential use against their owners.

The investigation highlighted the case of a teenager who turned to ChatGPT to inquire about his father’s discussion of a potential financial settlement, before his private conversations appeared in public records after he filed a lawsuit against Meta, Snapchat, TikTok, and YouTube. His lawyers later confirmed that the conversations did not affect the outcome of the case.

The investigation also revealed that conversations could reach authorities through the AI companies themselves. OpenAI reported to the Federal Bureau of Investigation a user in Florida who repeatedly told ChatGPT that he planned to assault and kill his former girlfriend. The man was subsequently arrested, pleaded guilty, and was sentenced to eight years of supervised release. OpenAI stated that it uses software to monitor risk indicators and refers concerning cases to human reviewers, and may notify law enforcement when a threat is imminent and credible.

These cases come amid a rapid increase in requests from government agencies and law enforcement for user data. In the second half of 2025, AI conversations currently lack legal protections comparable to attorney-client or doctor-patient privilege. Sam Altman, CEO of OpenAI, has called for specific legal protections for these conversations, as experts anticipate their growing presence in investigations and lawsuits as AI usage expands.