Federal court rules client use of generative AI not protected by attorney-client or work product privilege

A federal court ruled that a criminal defendant's written exchanges with a publicly available generative AI platform are not protected by attorney-client privilege or work product privilege. The U.S. District Court for the Southern District of New York issued the decision in United States v. Bradley Heppner. The defendant used the platform on his own initiative without instruction from counsel. He prepared documents outlining potential defense strategy and responses to charges after receiving a grand jury subpoena. The court applied the standard privilege test. It found no attorney-client relationship with the AI platform because the platform is not a lawyer. The platform's privacy policy stated that it collects inputs and outputs, uses them for training, and may disclose them to governmental authorities. This eliminated any reasonable expectation of confidentiality. The court also held that the defendant's later sharing of the materials with counsel did not make them privileged. The documents did not qualify as work product. They were prepared without direction from counsel and did not reflect counsel's strategy. The ruling confirms that established rules of privilege apply to generative AI tools.