Quick Buzz Feed
In February 2026, two federal courts issued rulings that together illustrate the unsettled landscape of privilege and work-product protection for AI-generated materials in litigation.
This section discusses the United States v. Heppner case, where a criminal defendant used Anthropic's Claude AI (consumer version) to prepare potential defense strategies, incorporating information from attorney discussions. Judge Rakoff ruled that these documents were not protected by attorney-client privilege because Claude is not a human attorney, its privacy policy allowed data collection and disclosure to third parties, and the defendant did not communicate with it for legal advice. Similarly, work-product protection was denied as the materials were not prepared under counsel's direction, though the court suggested a different outcome if counsel had directed the AI use.
This section details the Warner v. Gilbarco, Inc. case involving a pro se plaintiff using ChatGPT in an employment discrimination lawsuit. Magistrate Judge Patti denied the defendants' motion to compel discovery into the AI use, concluding that the AI-assisted litigation materials were protected by the work-product doctrine. The court deemed ChatGPT and similar generative AI programs as 'tools, not persons,' and rejected the argument that using AI constituted disclosure to a third party that would waive protection. The request for discovery was characterized as a 'fishing expedition' lacking relevance under Rule 26.
The article reconciles the seemingly conflicting outcomes of the Heppner and Warner cases by highlighting crucial factual differences. Key takeaways for litigants include: the importance of selecting AI tools with robust contractual confidentiality protections (enterprise vs. consumer platforms), the critical role of attorney direction in AI usage to preserve privilege, and the necessity for organizations to update AI governance policies to mitigate risks, such as restricting privileged information input into consumer platforms. Litigants should also anticipate and prepare for discovery requests concerning AI usage, either to defend against them or strategically pursue them. The decisions mark an early chapter in applying traditional privilege principles to evolving AI technology, with the contractual and technical features of AI platforms and attorney involvement being pivotal factors.