A federal judge recently concluded that the defendant in a white-collar securities dispute may not claim that his conversations with the artificial intelligence (“AI”) tool, Claude, are privileged. Litigators and clients now must take heed.

In United States v. Heppner, U.S. District Judge Jed S. Rakoff of the U.S. District Court for the Southern District of New York issued a written memorandum on February 17, 2026, opining that neither the attorney-client privilege nor the work-product doctrine protected documents generated after receipt of a grand jury subpoena by defendant Bradley Heppner using Claude, a generative AI tool, from government inspection.

This ruling, in a case of first impression, may have profound nationwide implications for parties seeking to withhold information from discovery in government investigations and litigation disputes.

In Heppner, the defendant was charged with securities fraud, wire fraud, conspiracy to commit securities fraud and wire fraud, making false statements to auditors, and falsifying corporate records. In the course of its investigation, the government seized “numerous documents and electronic devices,” including thirty-one documents memorializing communications between Heppner and Claude. The AI-generated documents explored possible charges, defense strategies, and arguments. Heppner’s counsel had not instructed him to run these searches through Claude or to prepare these kinds of reports.

The government moved for a ruling that the documents were not protected by either the attorney-client privilege or the work-product doctrine. Defense counsel, meanwhile, argued that the AI-generated documents were privileged because (1) they were prepared using information that the defendant had learned from his attorneys; (2) the defendant had created the documents in preparation for speaking with his counsel; and (3) the defendant had, in fact, subsequently shared the AI-generated documents with counsel.

In his written memorandum, Judge Rakoff noted that the attorney-client privilege protects from disclosure communications (1) between a client and his or her attorney (2) that are intended to be, and in fact were, kept confidential (3) for the purpose of obtaining and providing legal advice. The judge concluded that the AI-generated documents at issue failed for at least two, and possibly, all three elements of the privilege analysis.

First, Judge Rakoff found that the defendant’s AI-generated documents were not in fact communications between the defendant and his counsel, rendering the attorney-client privilege inapplicable. Specifically, the Court compared the defendant’s use of Claude to the defendant’s use of cloud-based internet software. The Court reasoned that a defendant’s use of cloud-based internet software would not trigger the protections of privilege unless counsel was in some way involved in the preparation of resulting documents. The Court applied the same logic to AI tools, concluding that privilege does not apply where counsel was not involved in the generation of the AI documents.

Second, Judge Rakoff found that the communications in the AI documents were not confidential because Claude’s written privacy policy, to which users of Claude consent, explicitly provides that:

Anthropic collects data on both users’ “inputs” and Claude’s “outputs,” that it uses such data to “train” Claude, and that Anthropic reserves the right to disclose such data to a host of “third parties,” including “governmental regulatory authorities.”

The Court determined that this “policy clearly puts Claude’s users on notice that Anthropic, even in the absence of a subpoena, may “disclose personal data to third parties in connection with claims, disputes[,] or litigation.

Third, Judge Rakoff found that the defendant did not communicate with the AI tool for the express purpose of obtaining legal advice or communicating with counsel. The Court determined that “the communications between Heppner and Claude were not privileged at the time they took place . . . because Heppner communicated with Claude of his own volition,” and not at the direction of legal counsel for purposes of obtaining legal advice from Claude. Further, the Court noted that Claude expressly disclaims that it is a lawyer and states that it does not have the ability to give legal advice. Thus, the Court determined that a privilege could not be triggered when communicating with a tool that explicitly disclaims that it is either an attorney or provides legal advice.

Finally, Judge Rakoff concluded that the AI-generated documents did not merit protection under the work-product doctrine because the AI documents were not prepared at the request of counsel and did not disclose counsel’s strategy.

Takeaways

As courts grapple with society’s rapid adoption of AI by applying traditional concepts to new technologies, it is clear that cases involving a party’s use of AI will turn on the facts specific to each case. Attorneys and clients alike should proceed cautiously with respect to AI use to guard against arguments that AI use waived the attorney-client privilege or attorney work product doctrine. Attorneys and clients should note the following:

  • If an AI provider reserves rights to retain, train on, or disclose user inputs, courts may find confidentiality and privilege was compromised.
  • In criminal actions, AI prompts and outputs relating to one’s own criminal investigation may not be protected after being seized by the government pursuant to a search warrant.
  • In civil actions, parties who independently use publicly available AI tools without their attorneys’ oversight to research their civil claims and/or defenses may similarly find their AI inputs and outputs are not protected from disclosure.
  • The ruling in United States v. Heppner leaves open whether counsel-directed enterprise AI use on a secure platform with strong confidentiality terms may be treated differently than a general Claude account.

If you have questions on issues relating to attorney client privilege and work product doctrine, specifically as they relate to AI, please reach out to the authors.

Epstein Becker Green Staff Attorney Ann W. Parks contributed to the preparation of this post.

Back to Commercial Litigation Update Blog

Search This Blog

Blog Editors

Authors

Related Services

Topics

Archives

Jump to Page

Subscribe

Sign up to receive an email notification when new Commercial Litigation Update posts are published:

Privacy Preference Center

When you visit any website, it may store or retrieve information on your browser, mostly in the form of cookies. This information might be about you, your preferences or your device and is mostly used to make the site work as you expect it to. The information does not usually directly identify you, but it can give you a more personalized web experience. Because we respect your right to privacy, you can choose not to allow some types of cookies. Click on the different category headings to find out more and change our default settings. However, blocking some types of cookies may impact your experience of the site and the services we are able to offer.

Strictly Necessary Cookies

These cookies are necessary for the website to function and cannot be switched off in our systems. They are usually only set in response to actions made by you which amount to a request for services, such as setting your privacy preferences, logging in or filling in forms. You can set your browser to block or alert you about these cookies, but some parts of the site will not then work. These cookies do not store any personally identifiable information.

Performance Cookies

These cookies allow us to count visits and traffic sources so we can measure and improve the performance of our site. They help us to know which pages are the most and least popular and see how visitors move around the site. All information these cookies collect is aggregated and therefore anonymous. If you do not allow these cookies we will not know when you have visited our site, and will not be able to monitor its performance.