Technology
AI chatbot logs are increasingly used as evidence in U.S. court cases
Law enforcement requests and court filings show growing use of user interactions with AI tools, while courts reject claims that the chats are privileged.
The short version
- A review of court records found chatbot logs cited in 12 U.S. criminal and civil cases over the last two years.
- OpenAI disclosed data on over 80 accounts following government requests in late 2025, more than four times the figure from the same period a year earlier.
- Judges currently do not recognize attorney-client privilege or similar legal protections for interactions with AI chatbots.
Key facts
- A review of public records and local news by The Washington Post identified chatbot logs in 12 court cases over the past two years.[Slashdot]
- OpenAI reports users to law enforcement if automated scanning and human reviewers determine there is an imminent and credible risk of harm.[Slashdot]
- In the second half of 2025, OpenAI complied with government requests for data from more than 80 accounts, a fourfold increase over the prior year's corresponding period.[Slashdot]
- A federal judge ruled against a fraud defendant, finding that legal privilege did not apply to queries made to the chatbot Claude.[Slashdot]
- OpenAI CEO Sam Altman has advocated for legal protections for chatbot interactions similar to attorney-client privilege, but current law does not recognize them.[Slashdot]
What remains uncertain
- The total number of police investigations and civil cases that draw upon chatbot records remains unknown because non-presented evidence is not fully cataloged in court filings.[Slashdot]
- OpenAI declined to state the total number of times it has proactively reported users to law enforcement.[Slashdot]