AI and the Attorney-Client Privilege
The Growing Use of AI in Legal Services
Like most professions, the practice of law has experienced an exponential increase in the use of generative AI. The application of AI to legal work has raised the question of whether AI-generated legal advice is protected by the attorney-client privilege.
Recent Federal Court Decisions on AI and Attorney-Client Privilege
Two U.S. federal court decisions, issued one week apart in February, reached opposite conclusions regarding the applicability of the privilege to AI.
In Warner v. Gilbarco, a Michigan federal district court denied a motion to compel production of AI-generated documents, ruling that the attorney work product protection applied because the litigant's use of an AI chatbot did not amount to disclosure to an adversary or in a manner that made it likely the information would reach an adversary.
On the other hand, in United States v. Heppner, a federal New York court ordered a criminal defendant to produce documents he had generated using an AI chatbot while seeking legal advice.
What These Decisions Mean for AI-Generated Legal Communications
The diverging decisions underscore how the legal framework is still unclear as to how the attorney-client privilege applies to AI. More case law is coming, undoubtedly. However, the court's opinion in Heppner gives us some insight.
The Heppner court held in part that litigants cannot expect confidentiality when using a public AI chatbot because such platforms collect data from all user inputs and outputs and reserve the right to disclose such data to third parties.
The court added, however, that had the defendant's attorney directed him to use AI, this might be akin to an attorney working with other professionals, such as an accountant or expert consultant, whose advice could fall within the protection of the attorney-client privilege.
The reasoning in Heppner suggests that attorney-directed use of an enterprise-grade AI platform (as opposed to a client using a public AI chatbot on their own) might mean that the resulting AI-generated communication is more likely to be privileged.
MacDonald Illig's Guidance on AI Confidentiality and Client Information
At MacDonald Illig, we advise our clients to refrain from using public AI platforms, even in a transactional setting, as the communications generated by public chatbots might become subject to disclosure to the other side should a dispute arise and litigation ensue.
Our firm uses only enterprise-grade AI platforms that operate under strict security protocols, including prohibitions on the use of client data for training purposes and restrictions on data retention. Such safeguards are designed to protect our clients' confidentiality and help ensure that AI-generated materials remain protected.
Questions About AI and Attorney-Client Privilege?
We continue to monitor developments in the law about the use of AI in legal services and constantly assess our enterprise-grade AI platforms to ensure the utmost in confidentiality.
For questions about how the use of AI may affect you in the legal context, please contact a MacDonald Illig attorney at 814-870-7600.
Legal Advice Disclaimer: The information presented on this website serves solely as general guidance and should not be construed as legal advice by MacDonald, Illig, Jones & Britton LLP as a replacement for seeking personalized legal counsel from a qualified attorney. MacDonald, Illig, Jones & Britton LLP does not assume liability for the accuracy or reliability of content hosted on any third-party websites accessible through links provided on this site.