Court ruling confirms AI chatbots lack legal privilege, prompting urgent warnings from US lawyers

AI Chatbots Face Legal Scrutiny After Landmark Court Ruling

A February 2026 court ruling has sent shockwaves through the legal community, forcing lawyers across the United States to reconsider how their clients interact with AI chatbots. US District Judge Jed Rakoff in New York delivered what legal observers describe as the first decision of its kind, ruling that conversations with AI chatbots are not protected by attorney-client privilege and can be seized as evidence in criminal and civil cases1

1

. The decision centered on Bradley Heppner, former chair of bankrupt financial services company GWG Holdings, who faces securities fraud charges and had used Anthropic's Claude to prepare reports about his case.

Judge Rakoff ordered Heppner to hand over 31 documents generated through his interactions with Claude, stating unequivocally that no attorney-client relationship exists "or could exist, between an AI user and a platform such as Claude"1

1

. The ruling hinged on three critical points: AI chatbots are not licensed attorneys, they owe no duty of confidentiality, and their terms of service explicitly permit data sharing with third parties including government authorities.

Major Law Firms Issue Urgent Client Advisories

The immediate aftermath of the court ruling has been a cascade of warnings from US lawyers. More than a dozen major law firms have issued client advisories cautioning against using AI tools for legal advice or sharing sensitive case information with platforms like ChatGPT and Claude1

1

. "We are telling our clients: You should proceed with caution here," said Alexandria Gutiérrez Swette, a lawyer at New York-based law firm Kobre & Kim4

4

.

Some firms have gone beyond advisories and embedded protective language directly into their client contracts. New York-based firm Sher Tremonte now includes explicit warnings in engagement agreements stating that sharing a lawyer's advice or communications with a chatbot could constitute a waiver of attorney-client privilege5

5

. This contractual approach reflects the legal profession's recognition that conversations with AI chatbots carry a significant risk of disclosure that could undermine client defenses.

Privacy Policies Expose Users to Legal Jeopardy

A closer examination of AI platform terms reveals why these warnings from US lawyers have become so urgent. Both OpenAI and Anthropic state in their privacy and usage terms that they can share user data with third parties, and both explicitly disclaim providing legal advice4

4

. During a February hearing, Judge Rakoff noted that Claude "expressly provided that users have no expectation of privacy in their inputs"1

1

.

When Rakoff asked Claude directly whether it could provide legal advice, the chatbot responded: "I'm not a lawyer and can't provide formal legal advice or recommendations," and recommended consulting "a qualified attorney who can properly assess your specific circumstances"3

3

. This self-disclaimer became part of the judicial reasoning that AI chats not privileged status should apply universally to public AI platforms.

Conflicting Rulings Create Uncertainty for Self-Represented Litigants

On the same day Rakoff issued his ruling, US Magistrate Judge Anthony Patti in Michigan reached what appeared to be a contradictory conclusion in Warner v. Gilbarco, Inc. Patti ruled that a woman representing herself in an employment discrimination lawsuit did not have to hand over her ChatGPT conversations, treating them as protected work-product rather than conversations with a third party5

5

. Patti wrote that ChatGPT and other generative AI programs "are tools, not persons".

Legal analysts note these cases are factually distinguishable. The work-product doctrine protections applied in Michigan involved self-represented plaintiffs in civil cases, while Heppner was a represented criminal defendant who used AI without attorney guidance. A third case, Morgan v. V2X in Colorado in March 2026, reached a similar protective conclusion for another pro se litigant. However, legal experts caution that these rulings don't establish broad protections and emphasize that anyone with legal representation should avoid using public AI platforms for case-related matters.

Implications Extend Beyond the Courtroom

The broader implications of this court ruling reach far beyond individual legal cases. As people increasingly rely on AI chatbots for various forms of advice, the Heppner decision establishes a precedent that could expose users to unexpected legal jeopardy3

3

. The ruling represents what some observers call "a new frontier for tech industry compliance with government demands," particularly given how many people have shared sensitive personal information with these platforms over recent years3

3

.

Law firms are now advising clients to use only "closed" AI systems designed for corporate use, which may provide stronger confidentiality protections, though even these remain largely untested in court5

5

. The guidance ranges from selecting AI platforms carefully to suggesting specific language for chatbot prompts that might preserve some level of protection. Legal professionals expect courts will continue grappling with how artificial intelligence intersects with bedrock legal protections governing attorney-client communications and materials prepared for litigation, making this an area where users should watch for evolving standards and additional case law that could either strengthen or further erode privacy expectations when using AI tools for legal advice.