AI Chats Are Not Confidential - Courts

A federal case in New York has broken new ground regarding AI. A CEO who was charged with fraud used “Claud,” an AI program, to research his own case. The FBI confiscated his computer and found those chats on it. The individual’s attorneys tried to restrict the access of the FBI on the basis of attorney-client privilege, but the court ruled that “chatting with an AI platform does not create a confidential attorney-client relationship.” 

It is pointed out that this could extend to the employment arena: if an HR team is using ChatGPT to figure out how to handle a performance issue, those conversations could be obtained by an opposing party in discovery. The only possible exception might be if a client’s attorney directs that client to conduct research using AI for the attorney’s work product – that might be protected; but no court has ruled on that. (It is noted that “incognito mode” in AI is not a protection; those conversations are still not private.)