Most tech industry products are easily accessible by the US government. Ring Doorbells have given the Los Angeles Police Department warrantless access to their customer’s camera footage. The FBI can extract your iPhones metadata to peep the content of your Signal messages. Google will happily comply with administrative subpoenas issued by Department of Homeland Security apparatchiks.
A new ruling by a New York federal judge now definitively includes AI chatbots in that list. As part of a protracted legal battle involving Brad Heppner, former chair of financial service company GWG Holdings, US District Judge Jed Rakoff ruled that AI chatbots aren’t subject to attorney-client privilege.
Maybe that sounds like a no-brainer, but evidently some people need the reminder. In preparing background materials for his attorneys, Heppner made the wise decision to enter various reports into Anthropic’s flagship chatbot Claude. The AI then spat out the preliminary reports, which his lawyers used to prepare his defense related to charges of securities and wire fraud, as Reuters reported.
The problem is that while attorney-client privilege protects most information exchanged between Heppner and his lawyers, it doesn’t extend to anything he jammed into Claude. As a result, the embattled financier is being forced to hand over 31 documents generated by Claude to the court.
In his opinion, Rakoff wrote that no attorney-client relationship exists “or could exist, between an AI user and a platform such as Claude.” The judge went even further, absolving the chatbot from any impropriety, because “Claude disclaims providing legal advice.”
“Indeed, when the Government asked Claude whether it could give legal advice, it responded that ‘I’m not a lawyer and can’t provide formal legal advice or recommendations’ and went on to recommend that a user ‘should consult with a qualified attorney who can properly assess your specific circumstances,'” Rakoff further explained.
Heppner’s alleged fraud notwithstanding, the ruling has major implications for AI chatbot users who could now be incriminating themselves, whether they know it or not.
The finding is already reverberating around law offices, according to Reuters. The white-collar defense firm Sher Tremonte, for example, updated its contract to reflect that “disclosure of privileged communications to a third-party AI platform may constitute a waiver of the attorney-client privilege.”
Again, tech companies handing personal data to Uncle Sam — willingly or otherwise — is nothing new. But when you consider how many people have dumped their entire brains into these chatbots over the past few years, it’s clear this ruling represents a new frontier for tech industry compliance with government demands.
More on AI and privacy: Woman Sues OpenAI, Saying ChatGPT Unleashed a Vicious Stalker Against Her and Did Nothing When She Begged for Help