US Courts Split on Privilege for AI-Generated Legal Work
Recent US court rulings conflict on whether AI-generated legal documents get attorney-client privilege.
Why it matters: Legal professionals need clarity on how using AI tools affects confidentiality and privilege risks. These rulings directly impact how law firms and corporate legal teams must handle AI-generated materials to safeguard client protections and litigation strategies.
- Feb 17, 2026: Southern District of NY ruled AI-generated documents lack privilege in United States v. Heppner.
- Heppner court emphasized no reasonable privacy expectation with public AI platforms due to their terms of service.
- Federal courts in Warner v. Gilbarco and Morgan v. V2X ruled AI tools are instruments, not third parties, allowing work product protection.
- The split rulings create immediate risks for companies using AI in legal research and litigation preparation.
The use of AI in legal workflows is under scrutiny as courts diverge on whether AI-generated materials qualify for attorney-client privilege and work product protection. In United States v. Heppner, the Southern District of New York ruled on February 17, 2026, that documents created using a publicly available AI platform were not protected by privilege or work product doctrines.
The Heppner court reasoned that communications with the AI did not constitute communications with an attorney. Furthermore, the AI platform's terms of service offered no confidentiality guarantees, negating any reasonable expectation of privacy.
Conversely, other federal courts took a different stance. In Warner v. Gilbarco, Inc. and Morgan v. V2X Inc., courts held that AI tools function as instruments rather than third parties. Therefore, their use does not automatically waive work product protection.
This conflict creates a fragmented legal landscape for attorneys and corporate legal departments that rely on AI for research or litigation preparation. The inconsistency raises confidentiality and privilege risks, potentially exposing sensitive materials during discovery or litigation.
Judge Hala Jarbou of the Western District of Michigan noted, "It should be obvious that any attorney who uses AI must scrupulously review its work product to ensure that the cited cases exist and that the citations accurately and fairly represent the underlying case law." This underscores caution in integrating AI outputs into legal documents.
By the numbers:
- February 17, 2026 — date of the pivotal Heppner ruling
- 31 — number of AI-generated documents evaluated in the Heppner case
- April 29, 2026 — date of Mintz analysis highlighting conflicting court views
Yes, but: Though Heppner reflects a strict view denying privilege, other courts recognize AI as a mere tool, preserving protections under some circumstances.
What's next: Further judicial decisions are expected to clarify how attorney-client privilege and work product protections apply to AI-generated materials across jurisdictions.