New Generative AI Court Ruling Puts HR on Notice: What to Know Now

A federal judge in New York ruled from the bench that documents an individual created using generative AI and sent to his lawyer were not protected by attorney-client privilege. The decision appears to be the first federal ruling of its kind on this issue. 

This ruling raises potential implications for HR teams that oversee internal investigations, employee communications, and workplace rules governing the use of generative AI.

Here’s how the case unfolded – and what it means for employers. 

Generative AI Ruling Stems From Fraud Charges

In November 2025, Bradley Heppner was arrested on securities and wire fraud charges. While serving as Board Chairman at GWG Holdings, Inc., Heppner allegedly misappropriated more than $150 million. 

“Heppner allegedly falsified documents, made misleading statements to investors and auditors, and obstructed an investigation by regulatory authorities. GWG’s subsequent bankruptcy resulted in over $1 billion in losses to retail investors,” said FBI Assistant Director in Charge Christopher G. Raia. 

The trial is scheduled to start on April 2, 2026.

How AI Became an Early Focal Point in the Case

After receiving a grand jury subpoena, Heppner used an AI tool to outline his defense strategy and potential legal arguments. He later shared that AI-generated material with his attorney.

When Heppner was arrested, federal agents executed a search warrant at his home, seizing electronic devices that contained approximately 31 AI-generated documents. 

Heppner’s attorney insisted that the documents were protected by attorney-client privilege.

The government filed a motion arguing that they were not, and the court agreed, noting three key points.

First, the communications between Heppner and the AI platform were between two non-attorneys, so they weren’t privileged from the start.

Second, the AI platform’s written privacy policy explicitly warned users that their input and outputs would train the model and could be shared with third parties. In the court’s view, this eliminated any “reasonable expectation of confidentiality,” waiving privilege.

And third, the court pointed out that Heppner was not instructed by his attorney to conduct generative AI searches. “Because Heppner communicated with [the AI platform] of his own volition,” what matters for attorney-client privilege is whether he intended to obtain legal advice from AI – not whether he shared AI’s output with his attorney, the court explained. 

Importantly, the AI platform includes a disclaimer that it cannot give legal advice. In fact, in this case, the government asked the AI platform if it could give legal advice. It replied: “I’m not a lawyer and can’t provide formal legal advice or recommendations.” It also recommended consulting with “a qualified attorney who can properly assess your specific circumstances,” the court noted.

In light of these three elements, the court determined the AI-generated documents were not protected by attorney-client privilege.

While Heppner’s situation is extreme, it highlights risks HR can encounter when using AI in internal processes that are later shared with counsel. Admittedly, few HR leaders would rely on AI for something as sensitive as a criminal defense, yet many already use generative tools for internal investigations or documentation – areas where privilege can apply.

Employment Attorney Weighs In 

When it comes to concerns about artificial intelligence, data accuracy and bias have received the lion’s share of the attention – at least so far. Now this generative AI ruling highlights new concerns: confidentiality and attorney-client privilege.

So what does this mean for HR teams? According to Richard Warren, Detroit shareholder with Ogletree Deakins, HR teams should focus on:

Disclosure risks: “Now that this decision is public, we should expect litigants, and perhaps state and federal agencies, to ask companies to produce generative AI prompts and results carried out by HR teams and managers,” Warren says. “Unlike a simple web search, generative AI prompts and responses can potentially reveal much more about the strengths and weaknesses of the employer’s position and could contain admissions used against the company in litigation.” 

Recommended controls: “As a result, generative AI-based legal research and analysis should only be done in a manner consistent with the attorney-client privilege, work-product doctrine and their requirements,” he advises.   

HR Takeaways

“The Heppner decision held that a non-attorney’s generative AI searches that were not performed at the direction of a lawyer were discoverable in litigation and were not covered by the attorney-client privilege or work product doctrine,” Warren notes.

Other courts could follow this reasoning, he points out, meaning HR professionals using generative AI to evaluate internal complaints — like prompting an AI tool to assess a harassment investigation — could find their prompts and responses subject to disclosure to opposing parties in litigation.

Action steps for HR:

• Reinforce data privacy protections. Many AI platforms store or review inputs to train their models, so assume anything entered could be accessible to third parties.
• Prohibit unsupervised AI use for legal research. Independent use of generative AI for legal or investigatory purposes can waive attorney-client privilege. Always loop in legal counsel first.
• Update policies and training programs. Embed AI-specific guidance into confidentiality and data security policies, using real scenarios in training.
• Coordinate with Legal and IT. Work jointly to vet AI tools, establish secure data practices, and define safe boundaries for employee use.

This ruling makes clear: AI-generated materials can undermine attorney-client privilege. Treat AI as you would any outside party when legal confidentiality is at stake. 

United States v. Heppner, No. 25 Cr. 503 (S.D.N.Y. 2/17/26).