The rapid adoption of Artificial Intelligence (AI) is reshaping how we communicate. While employees may be tempted to turn to AI to gather information and for assistance drafting workplace grievances, they must approach the use of AI with caution. The use of AI in legal proceedings is quickly becoming a battleground in employment litigation. Courts will increasingly be asked to address whether and to what extent AI chat histories can be used during litigation, and two recent federal court judges reached nearly opposite conclusions on the matter.

Understanding the Limitations on the Discovery of AI Searches in Employment Litigation

During the discovery phase of a lawsuit, the parties collect and exchange information to help the attorneys and the judge to better understand the facts of the case. However, discovery is limited to information that is relevant, proportional to the needs of the case, not privileged, and not prepared in anticipation of litigation. Notably:

• The attorney-client privilege shields communications between a client and their attorney from discovery.
• The work product doctrine protects from discovery documents and other materials prepared in anticipation of litigation.

Courts are increasingly being asked to address whether AI-generated content, such as meeting summaries, emails, and chat histories used to generate grievances or respond to disciplinary matters, is subject to discovery.

Two Federal Court Cases, Two Different Results

In February 2026, judges in two different federal court cases reached nearly opposite conclusions on the discovery of AI-generated information in litigation.

United States v. Heppner - No Privilege or Protection for AI-Generated Content

In Heppner, a criminal defendant in New York was indicted on various fraud-related charges. During the arrest and search of his home, FBI agents seized documents and electronic devices that included Heppner’s communications with Claude, a generative AI platform operated by Anthropic, which he later shared with his attorneys. The defendant claimed the information was privileged, as it was prepared for the purpose of speaking with counsel to obtain legal advice. The judge ruled that the communications with Claude were not privileged because:

1. They were not between Heppner and his legal counsel.
2. They were not confidential.
3. They were not created for the purpose of obtaining legal advice, as Heppner’s lawyer did not direct him to communicate with Claude (although the court noted that had this been the case, it would have been a closer call).

The court also ruled that the work product doctrine did not apply because, even if Heppner were communicating with Claude in anticipation of litigation, he was not doing so at his counsel’s direction.

Warner v. Gilbarco, Inc. - Chat GPT Interactions Are Protected Work Product

In Warner, a Michigan woman filed a pro se lawsuit against her employer in 2024, alleging gender, race, and national origin discrimination, and retaliation under state and federal law. Her employer asked the court to order her to produce “all documents and information concerning her use of third-party AI tools in connection with this lawsuit,” including, for example, her prompts/queries and the AI outputs.

The company claimed the ChatGPT materials were relevant, discoverable, and not protected by the work product doctrine because the plaintiff is self-represented and voluntarily disclosed the information to ChatGPT, a third party.

The plaintiff claimed the discovery request was seeking her “internal analysis and mental impressions – i.e., her thought process – rather than any existing document or evidence, which is not discoverable as a matter of law.”

The court ruled in favor of the plaintiff, finding that the ChatGPT materials were not discoverable because:

1. They were not relevant;
2. Even if they were, the request was not proportional; and
3. They were protected by the work product doctrine.

The Takeaway

While these cases reflect two federal court judges’ initial forays into the discovery of AI-generated material, a split appears to be emerging over whether interactions with AI are protected by the attorney-client privilege or the work product doctrine. As a result, employees must take caution in their interactions with AI tools, particularly when they are used to investigate potential claims, prepare grievances, or respond to disciplinary investigations. The best course of action is to consult with an experienced employment lawyer who can provide guidance, within the scope of the attorney-client privilege, about the best course of action – including the best approach to whether, and how, AI should be used in connection with the claims against your employer.

Contact the Employment Lawyers at Madsen, Prestley & Parenteau, LLC, Today

With offices in Hartford and New London, the attorneys at Madsen, Prestley & Parenteau, LLC, represent individuals in employment law matters throughout Connecticut. To learn more and discuss how we can assist you, contact us online or call 860-246-2466 to schedule a confidential consultation.