AI-Generated Documents lack Privilege
- Gregory Stanley
- 6 hours ago
- 3 min read
The case of United States v. Heppner has brought a new and critical issue to the forefront of legal practice: how courts treat documents generated by artificial intelligence tools. Bradley Heppner, the former chairman of GWG Holdings, Inc., faces serious charges including securities fraud and conspiracy. A key development in this case involves 31 documents created using a commercial AI tool called Claude, which prosecutors seized during a search of Heppner’s residence. The court’s decision that these AI-generated documents are not protected by attorney-client privilege or the work-product doctrine has significant consequences for how AI tools are used in legal settings.
This post explores the details of the Heppner case, the court’s ruling on AI-generated documents, and why this matters for clients, lawyers, and anyone using AI in legal environments.
Background of United States v. Heppner
Bradley Heppner was charged by federal prosecutors in the Southern District of New York with multiple offenses related to alleged financial misconduct at GWG Holdings, Inc. The government claims Heppner diverted corporate funds through a related entity while hiding these actions from auditors and investors. The trial is scheduled for April 2026 before Judge Jed S. Rakoff.
During the investigation, authorities seized 31 documents created using Claude, an AI-powered text generation tool. Heppner argued these documents were protected by attorney-client privilege because he shared them with his lawyers while preparing his defense. The government disagreed and asked the court to rule that these materials could be used as evidence.
The Court’s Decision on AI-Generated Documents
On February 10, 2026, Judge Rakoff ruled that the AI-generated documents are not protected by attorney-client privilege or the work-product doctrine. This decision rests on several key points:
AI is not a lawyer: Communications with AI tools do not count as communications with an attorney.
No confidentiality: Using a public AI platform is treated as sharing information with a third party, which breaks confidentiality.
Intent alone does not create privilege: Even if Heppner believed the documents would help his lawyers, that belief does not grant privilege.
Work product requires attorney involvement: Documents created by a client without direction from counsel do not qualify as protected work product.
Because of this ruling, prosecutors can review and use all 31 AI-generated documents in their case.
What This Means for Legal Practice and Clients
This ruling sends a clear message about the risks of using commercial AI tools for legal work or strategy. Here are the main takeaways:
AI Use Can Create Discoverable Evidence
When clients or lawyers use public AI platforms to draft notes, analyses, or strategy documents, those materials may be seized and used by regulators or opposing parties. This means:
No automatic protection: AI-generated content is not automatically confidential or privileged.
Risk of exposure: Sensitive legal strategies or internal thoughts created with AI tools can become evidence.
Careful tool choice: Using AI tools with strong privacy and security policies is crucial.
Privilege Requires Lawyer Involvement
Attorney-client privilege protects communications between a client and their lawyer. This case clarifies that:
Self-generated AI content is not privileged: Even if shared later with lawyers, documents created independently with AI do not gain protection.
Work product protection depends on counsel’s direction: Only materials prepared at the lawyer’s request or under their supervision qualify.
Clients must be cautious: Relying on AI tools without lawyer involvement risks losing privilege.
Practical Advice for Using AI in Legal Contexts
The Heppner ruling highlights the need for clear policies and awareness when using AI tools in legal matters. Here are some practical steps:
Avoid using public AI tools for confidential legal work: Use secure, private platforms designed for legal professionals.
Involve lawyers early: Have attorneys direct or supervise the creation of documents intended to be privileged.
Educate clients and staff: Make sure everyone understands that AI-generated content may not be protected.
Keep records of lawyer involvement: Document when and how lawyers participate in drafting or reviewing AI-generated materials.
Review AI tool terms of service: Understand how data is stored and shared by the AI provider.
The United States v. Heppner decision marks a turning point in how AI-generated documents are treated in the legal system. It underscores the importance of understanding the limits of privilege and confidentiality when using AI tools. For clients and lawyers, the key lesson is clear: AI can be a powerful aid, but it must be used carefully and with legal oversight to avoid unintended risks.
Comments