The consultation, use, or inclusion of Artificial Intelligence (A.I.) by a client into any part of the litigation process is strongly discouraged. As A.I. tools, such as ChatGPT, Claude, and Gemini, become increasingly accessible, it is imperative to understand the risks associated with providing third-party platforms with detailed information regarding your case.
When working with an attorney, the client is granted protection by the attorney-client privilege - a legal principle that protects confidential communications between the attorney and client. This protection covers all privileged communications relevant to the legal services provided by the attorney, including verbal discussions, written correspondence, emails, and text messages. However, there are few exceptions to the protection offered by this privilege; confidentiality may be compromised when privileged information is disclosed to a third party. For example, when a client enters information about their case into a third-party AI system, some jurisdictions have concluded that information is no longer protected by the attorney-client privilege and it becomes discoverable.
In the case of United States v. Heppner, the Defendant, Bradley Heppner, consulted with the generative AI platform “Claude” regarding defense strategy proposed by his counsel. United States v. Heppner, 828 F. Supp. 3d 292 (S.D.N.Y. 2026). The strategies included sensitive information that outlined what he might argue in response to his anticipated indictment. Despite attempts by Heppner’s counsel, the Court held that the information put into “Claude” lacked at least two elements of the attorney-client privilege: the conversation was not between Heppner and his counsel and they were not confidential as “Claude” is a third-party platform that collects data on both inputs and outputs from the user. Consequently, the information Heppner provided to “Claude” was admissible in court.
In turn, attorneys are facing increasing regulation to prevent the same practice of disclosing personal details to third-party platforms. Effective January 1st, 2027, California Senate Bill 574 requires attorneys to use appropriate technologies competently, supervise their use, verify material outputs, protect confidential information, and ensure that technology does not replace the lawyer’s judgment. While A.I. may offer reasonable tools for attorneys to utilize in their practice, such as researching relevant statutes and cases or drafting correspondence, the attorney must take reasonable actions to verify the accuracy of the product produced by the platform when using such tools. Additionally, if an attorney uses a generative A.I. platform, they may not provide personal or identifiable information.
A.I. users are not offered substantial privacy protections, when using their standard consumer services. Therefore, from the standpoint of preserving one’s privacy and privilege protections, the safest practice is to avoid putting facts, evidence, or legal inquiries into most consumer-oriented A.I. platforms.
