Private ChatGPT Conversations Used as Evidence in US Court Cases
Private conversations that users had with ChatGPT and Claude have been used as evidence in at least 12 US court cases over two years. Courts have found that AI chatbot communications do not have legal privilege comparable to attorney-client confidentiality, allowing prosecutors and opposing lawyers to obtain and use these conversations as evidence.
What happened
ChatGPT conversations have been cited as evidence in at least 12 criminal and civil court cases over the past two years, according to a Washington Post review of public records and local news reports.
In one case, a teenage boy identified as R.K.C. sued Meta, Snap, TikTok and YouTube alleging their platforms contributed to social media addiction and mental health problems; defense attorneys obtained ChatGPT conversations in which the teenager discussed deeply personal issues.
In a Florida case, a man identified as Darren Zhou allegedly spent months telling ChatGPT that he planned to rape and murder his ex-girlfriend; OpenAI flagged the conversations and alerted the FBI, which passed the information to local police.
A federal judge in New York rejected financial executive Bradley Heppner's argument that his conversations with Anthropic's Claude about legal defense strategies should be protected, finding that communicating with an AI chatbot was not equivalent to communicating with an attorney.
Reported impact
Users who confided personal information to chatbots discovered their conversations could become evidence and part of the public record.
Millions of people increasingly use AI to discuss personal matters under the assumption that conversations will remain private, but those conversations may not remain confidential once police or opposing lawyers obtain them.
Index analysis
Chatbot conversations can enter legal proceedings through civil discovery, legal subpoenas, court orders, or from a suspect's phone during criminal investigations when the owner consents to a search.
Ordinary conversations with AI chatbots do not currently have a generally applicable attorney-client privilege equivalent to communications with lawyers, as courts have found that chatbot communications do not qualify for such legal protection.
Governance controls
OpenAI has published procedures for responding to valid civil subpoenas and court orders seeking user information, saying it reviews such requests and produces information only when permitted under applicable law and will notify affected users before disclosure when legally possible and appropriate.
OpenAI uses automated systems and human review to detect potentially dangerous activity and may refer cases to law enforcement when reviewers determine there is a serious threat.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.