ChatGPT may feel like a private space to share thoughts, ask sensitive questions or seek advice. But recent court cases show that those conversations can sometimes become part of legal proceedings, raising serious questions about AI privacy.

ChatGPT chats used as evidence are becoming a real issue as AI assistants become part of everyday life. It was reported that ChatGPT conversations were cited in at least 12 criminal and civil cases over the past two years. One case involved a teenager identified in court as R.K.C., who sued Meta, Snap, TikTok and YouTube over alleged social media addiction and mental health problems. During the case, attorneys obtained ChatGPT conversations in which he discussed deeply personal matters, including something his father had told him.

The bigger concern is that does talking to AI have the same privacy protection as talking to a lawyer? Right now, the answer in the US is no.

A federal judge rejected financial executive Bradley Heppner’s argument that his conversations with Anthropic’s Claude about legal defense strategies should receive attorney-client protection. Heppner was later convicted of fraud charges linked to a scheme involving more than $150 million.

How can AI conversations enter a court case?

Chatbot records can be requested during civil discovery or through valid commands and court orders. OpenAI says it reviews government and legal requests and provides user information only when legally permitted. When possible, it also says affected users may be notified before disclosure. Police investigations can also involve chatbot data stored on a suspect’s device.

OpenAI CEO Sam Altman has acknowledged the risk, saying, “If you go talk to ChatGPT about your most sensitive stuff and then there’s like a lawsuit or whatever, we could be required to produce that.”

The issue is growing as government requests increase. During the second half of 2025, OpenAI reported receiving 224 non-content requests, 75 content requests and 10 emergency requests from government agencies.

What happens to sensitive information shared with AI?

Deleting a conversation does not mean every privacy concern disappears. OpenAI says deleted chats and Temporary Chats are generally removed within 30 days, subject to legal and security exceptions. Researchers have also warned about AI systems leaking or memorizing information. In 2025, chatbot transcripts from Anthropic’s Claude reportedly appeared in Google search results.

For users, there is one important lesson: do not assume an AI chatbot is a confidential counselor, lawyer or therapist.

The future could bring stronger AI privacy rules

Sam Altman has argued that AI conversations should eventually receive legal protections similar to certain professional communications. As Business Fortunes observes, AI becomes more deeply involved in personal and professional decisions, lawmakers may face growing pressure to define the boundaries of chatbot privacy.

 

FAQs

Can ChatGPT conversations be used in court?

Yes. Depending on the case and applicable law, authorities or legal parties may obtain chatbot records through lawful processes.

Are ChatGPT conversations protected like attorney-client communications?

No broad AI privilege currently exists under US law.

Can police obtain ChatGPT data?

Yes, if they follow applicable legal procedures, such as obtaining required legal orders or accessing data from a device during a lawful investigation.

Does deleting a ChatGPT chat remove it immediately?

Not necessarily. OpenAI says deleted and Temporary Chats are generally deleted within 30 days, with certain legal and security exceptions.

Will AI conversations get stronger legal protection?

Possibly. As AI use expands, lawmakers and courts may develop clearer rules around privacy, disclosure and AI-generated records.