by Eric D. Morton
The law is evolving regarding the use of artificial intelligence. There are dozens of major copyright lawsuits being fought in the courts. New issues keep appearing. One is whether or not any legal privilege exists in the use by a consumer of generative AI.
A recent federal decision highlights a risk for using generative AI tools. In United States v. Heppner, No. 25-cr-00503-JSR (S.D.N.Y. Feb. 17, 2026), the court held that materials created by a nonlawyer criminal defendant using a public, nonenterprise AI platform were not protected by the attorney-client privilege or the work product doctrine.
This ruling is among the first to squarely address how traditional privilege principles apply to an individuals independent use of generative AI, and it offers important guidance for businesses navigating AI adoption.
Background
Federal prosecutors charged Bradley Heppner with multiple criminal offenses arising from an alleged scheme to defraud investors of more than $150 million. During a search of his home, FBI agents seized documents reflecting Heppner’s use of a consumer AI platform in exploring defenses while he was under investigation.
Notably:
- Heppner used the AI tool on his own initiative, without direction from counsel.
- He used the tool to prepare reports analyzing potential defense strategies.
- He later shared the AI-generated materials with his attorneys.
After the government obtained these materials, Heppner argued they were protected by the attorney-client privilege and the work product doctrine.
The Court’s Holding
The court rejected those arguments, concluding that:
- No attorney-client privilege applied because the communications were not made between client and counsel for the purpose of obtaining legal advice.
- No work product protection applied because the materials were not prepared by or at the direction of an attorney, nor in anticipation of litigation under counsel’s guidance. Sharing later is not enough.
- The use of a public, nonenterprise AI tool further undermined any expectation of confidentiality. Confidentiality is critical.
The court treated the AI tool as a non-lawyer third party.
Practical Guidance for Businesses
This case has implications businesses and not just an individual under criminal investigation. Often businesses find themselves in disputes. Employees and officers involved in the dispute might seek a quick search of their potential rights, claims or defenses on a generative AI platform. None of those searches or queries, and the information entered by the employees or officers is privileged. None of the results are privileged either. Everything can be discovered in subsequent discovery during the course of a lawsuit.
We suggest companies develop policies for:
- Establishing clear AI usage policies for employees, particularly regarding legal matters.
- Restricting use of public AI tools for sensitive or litigation-related analysis.
- Using enterprise AI solutions with contractual confidentiality protections where appropriate.
- Requiring legal involvement before generating AI-assisted legal analyses.
- Training employees on privilege risks associated with AI tools.
Looking Ahead
As courts continue to confront the intersection of generative AI and longstanding legal doctrines, decisions like Heppner signal a cautious approach: traditional privilege rules will not be relaxed simply because new technologies are involved.
Organizations adopting AI should proactively align their practices with these principles to avoid unintended disclosure risks.
If you have questions about this decision or would like assistance developing AI governance policies, please contact us.
Eric D. Morton is the principal attorney at Clear Sky Law Group, P.C. He can be reached at 760-722-6582, 510-556-0367, and emorton@clearskylaw.com