Court Rules Use of Public AI Tools Can Waive Attorney-Client Privilege

Landmark decision in U.S. v. Heppner raises concerns over confidentiality when lawyers and clients use generative AI models like ChatGPT.

Mar. 30, 2026 at 8:28pm

Got story updates? Submit your updates here. ›

A highly detailed 3D illustration of a glowing, neon-lit cybersecurity server rack, representing the digital infrastructure that powers modern legal technology. The rack is illuminated by pulsing cyan and magenta lights, creating a sense of high-tech complexity and the risks of exposing sensitive legal data.As lawyers increasingly leverage AI tools, a landmark court ruling highlights the need to safeguard attorney-client privilege in the digital age.NYC Today

In a groundbreaking ruling, a U.S. District Court in New York has determined that a criminal defendant's use of public generative AI platforms like ChatGPT to discuss his defense strategy waived both attorney-client privilege and work product protections. The court found that because the AI platforms' terms of service allow for potential disclosure of user inputs and outputs, communications with these tools cannot be considered confidential, undermining the core tenets of privilege.

Why it matters

This decision highlights the growing risks that the use of public AI tools poses for the sacrosanct attorney-client relationship and the ability of lawyers to protect their work product. As more attorneys and clients turn to generative AI for efficiency and cost-savings, this ruling serves as a wake-up call about the need to carefully consider privilege and confidentiality implications.

The details

The case, U.S. v. Heppner, involved a corporate executive charged with securities fraud who used a public AI platform to research potential defense strategies before consulting his attorney. When the government later obtained the AI-generated reports, Heppner tried to shield them using attorney-client privilege and work product doctrine. However, the court ruled that because Heppner was not directed by counsel to use the AI tool, and because the platform's terms made clear the communications were not confidential, any claim of privilege had been waived.

  • In February 2026, the U.S. District Court for the Southern District of New York issued its ruling in U.S. v. Heppner.
  • Prior to his indictment, and after receiving a grand jury subpoena, Heppner used a public AI platform to research potential defense strategies.

The players

U.S. v. Heppner

A landmark court case that addressed whether a criminal defendant waived attorney-client privilege and work product protections by using a public generative AI platform to discuss his defense strategy.

Heppner

A corporate executive indicted on multiple counts of securities fraud who used a public AI platform to research potential defense strategies before consulting his attorney.

Got photos? Submit your photos here. ›

What’s next

Attorneys and their clients must carefully consider the confidentiality implications of using public generative AI platforms, and may need to secure informed consent or explore the use of closed, legally-trained AI tools to avoid potential privilege and work product issues.

The takeaway

The Heppner ruling serves as a stark reminder that the use of public AI tools, even for legal research and strategy, can inadvertently waive core legal protections like attorney-client privilege. Lawyers must proactively address this emerging risk with clients to maintain the confidentiality that is essential to the attorney-client relationship.