Site icon Jimerson Birr

Privilege in the Age of AI After United States v. Heppner

Privilege-in-the-Age-of-AI-After-United-States-v.-Heppner.

Privilege-in-the-Age-of-AI-After-United-States-v.-Heppner.

A recent decision from the Southern District of New York addresses a question that is no longer theoretical. Specifically, how traditional privilege doctrines apply when parties use generative artificial intelligence in connection with potential litigation. In United States v. Heppner, 2026 U.S. Dist. LEXIS 32697 (S.D.N.Y. Feb. 17, 2026), the court held that communications with a publicly available AI platform were not protected by either attorney-client privilege or the work product doctrine.

The opinion is framed as a question of first impression, but its significance lies in how firmly it applies existing doctrine. The court does not treat AI as a special case. It treats it as a modern version of a familiar problem, disclosure of information to a third-party.

Background

After receiving a grand jury subpoena and becoming aware he was a target of the investigation, the defendant used a publicly available AI platform to generate written materials analyzing potential defenses and legal strategy. Those materials were later seized pursuant to a search warrant. He asserted privilege on the grounds that the materials incorporated information obtained from counsel, were prepared in anticipation of litigation, and were ultimately shared with counsel. The court rejected both the attorney-client privilege and work product claims.

Attorney Client Privilege

Applying settled law, the court focused on whether the communications were between a client and counsel, intended to be confidential, and made for the purpose of obtaining legal advice.

They were not communications with an attorney. The court emphasized that discussions of legal issues between non-lawyers are not privileged, and that non-privileged materials do not become privileged simply because they are later provided to counsel.

They were not confidential. The court relied on the platform’s privacy policy, which disclosed that user inputs and outputs could be retained, used for training, and shared with third parties, including in connection with litigation. In that setting, there was no reasonable expectation of confidentiality.

They were not made for the purpose of obtaining legal advice. The defendant did not act at counsel’s direction, and the relevant inquiry was whether legal advice was sought from the recipient of the communication. Here, it was not.

The court also addressed the consequence of disclosure. To the extent the defendant input information learned from counsel into the AI platform, any privilege attached to that information was waived by sharing it with a third-party.

AI as a Third Party, Not a Tool

One of the more useful aspects of the opinion is how it handles the argument that AI should be treated like other software tools. The suggestion was that using an AI platform is no different than using a word processor or other technology, and therefore should not be treated as a disclosure.

The court declined to adopt that view. Even if the analogy were accepted, the use of software is not inherently privileged. Privilege depends on the existence of a protected relationship. As the court explained, recognized privileges require a trusting human relationship with a professional who owes fiduciary duties and is subject to regulation.

No such relationship exists with a public AI platform. At a minimum, using that platform involves disclosing information to a third-party that retains and may share it. That is enough to defeat any claim of confidentiality.

Work Product Doctrine

The work product doctrine did not apply for similar reasons. Although the materials were created after the defendant anticipated litigation, they were not prepared by or at the direction of counsel.

The defendant acted on his own initiative, and his counsel did not direct the use of the AI platform. The court also drew a distinction that will likely recur in future cases. Materials that affect counsel’s strategy are not the same as materials that reflect counsel’s strategy. Only the latter fall within the doctrine’s core protection.

The court left open the possibility that a different analysis might apply if the use of an AI system were directed by counsel in a manner analogous to an agent assisting in the provision of legal advice, but it did not resolve that issue.

Broader Implications

The decision reinforces a point that extends beyond AI. Privilege turns on relationship and confidentiality, not on the sophistication of the technology being used. When information is shared with a third-party outside a protected relationship, the default rule applies.

In that sense, the case is less about artificial intelligence and more about the continued application of traditional doctrine in new settings. AI simply makes the line easier to cross. What might previously have been a private draft or internal analysis can become a third-party communication with a few keystrokes.

The decision also has implications for how discovery disputes will develop. If communications with AI platforms are treated as ordinary third-party disclosures, there is little doctrinal basis to treat them differently from emails, texts, or other informal communications. Parties should expect that prompts, outputs, and related materials may be requested in discovery where they bear on claims, defenses, or the development of a party’s position.

Conclusion

Heppner confirms that courts are unlikely to carve out special treatment for AI-assisted work. The existing framework governs. Communications with a public AI platform are treated as disclosures to a third-party, and materials generated through that process do not become privileged simply because they relate to legal issues or are later shared with counsel.

As the use of these tools becomes more common, the practical question will not be whether the doctrine changes, but how often parties inadvertently step outside of it.

Questions About Privilege, AI, and Litigation Risk?

If you are facing a government investigation, navigating a discovery dispute, or have questions about how your use of AI tools may affect privilege protections, the attorneys at Jimerson Birr are here to help. Our team has deep experience advising clients on complex litigation strategy, privilege issues, and the evolving intersection of technology and the law. Contact us today at jimersonbirr.com or call our office to schedule a consultation. We’re ready to help you understand your rights and protect your interests.

Exit mobile version