NEWS & INSIGHTS

In United States v. Heppner, a Federal Court holds that communications with public Artificial Intelligence (“AI”) platforms are not privileged.

In a significant decision at the intersection of generative AI and legal privilege, the U.S. District Court for the Southern District of New York held that a criminal defendant’s communications with a publicly available AI platform were not protected by attorney-client privilege or the work product doctrine. The ruling underscores a critical point for companies, executives, and individuals: unsupervised use of public AI tools for legal strategy or analysis may jeopardize confidentiality protections.

The defendant in United States v. Heppner was charged in federal court with multiple fraud-related offenses. After retaining counsel, he independently used a public generative AI platform to generate written analyses, defense theories, and strategic considerations related to his case. The AI-generated materials were later shared with his attorneys. During the government’s investigation, those materials were seized pursuant to a warrant. The defense asserted attorney-client privilege and work product protection over the documents.

The court rejected those claims. 

The court concluded that the AI-generated communications were not privileged, reasoning as follows:

1. No Attorney-Client Communication

Attorney-client privilege protects confidential communications between a client and an attorney made for the purpose of obtaining legal advice. The court held that:

• Communications with a public AI platform are not communications with an attorney.
• An AI tool is not a licensed professional bound by duties of confidentiality.
• Accordingly, the threshold requirement for privilege was not satisfied.

2. No Reasonable Expectation of Confidentiality

Privilege requires that communications be made in confidence. The court emphasized that:

• Public AI platforms operate under terms of service that may permit data storage, review, or third-party access.
• Users of such platforms may lack a reasonable expectation that their inputs are confidential in the legal sense required for privilege protection.

3. Subsequent Disclosure to Counsel Does Not Cure the Defect

The defendant argued that sharing the AI-generated documents with counsel brought them within the privilege.

The court disagreed, holding that:

• Documents created independently outside the attorney-client relationship do not become privileged merely because they are later transmitted to counsel.
• Privilege does not attach retroactively.

4. Work Product Doctrine Inapplicable

The work product doctrine protects materials prepared by or for attorneys in anticipation of litigation. The court found that:

• The materials were generated by the defendant independently.
• They were not prepared at the direction of counsel.
• They were not attorney mental impressions or legal analysis.
• Therefore, work product protection did not apply.

Why This Decision Matters

Although the decision arises in a criminal context, its implications extend broadly to corporate and civil matters. Companies should assume that:

• Employees who input sensitive internal information into public AI systems may be creating discoverable material.
• Internal investigations, compliance analyses, or litigation strategy generated through public AI tools may not be protected.

The ruling signals that courts will:

• Apply traditional privilege doctrine strictly.
• Decline to extend privilege protections to communications with software platforms.
• Focus on confidentiality, attorney involvement, and reasonable expectations of privacy.

The court did not:

• Prohibit the use of AI in legal practice.
• Address secure enterprise AI platforms operating under negotiated confidentiality terms.
• Foreclose the possibility that attorney-directed AI use could qualify as work product under different facts.
• Future cases may refine these distinctions.

Conclusion

United States v. Heppner represents an early but influential judicial signal: traditional privilege principles remain technology-neutral. Communications with public AI platforms are not equivalent to communications with legal counsel.