Texas Judge Suggests Non-Lawyer AI ‘Chats’ May Be Protected Attorney Work Product

Quick Hits

• A Texas judge ruled that conversations with generative AI tools could be protected as attorney work product under Texas law, diverging from a recent landmark federal case.
• The decision implies that using AI does not automatically waive privilege, as such conversations may not be likely to be disclosed to adversaries.
• Employers may want to monitor employee interactions with AI to manage potential risks to attorney-client privilege and confidentiality.

On June 3, 2026, Judge Grant Dorfman for the Texas Business Court Eleventh Division issued a minute entry in Tate Group Automotive, LLC v. Legacy Automotive Capital, LLC, ruling that a non-lawyer’s AI conversations, prepared in anticipation of litigation, could qualify as privileged work product under Texas procedural rules.

Judge Dorfman indicated that AI conversations could be covered by the work-product protection under Texas law, citing the Texas rules of civil procedure that work product is “material prepared or mental impressions developed in anticipation of litigation or for trial by or for a party….” (Emphasis in decision.) In other words, work product is not limited to materials developed by lawyers and may include materials generated by the principal of a corporate entity—as was the case in Tate, or by pro se litigants.

In this vein, Judge Dorfman cited with approval two recent district court opinions involving the use of AI by self-represented litigants in employment litigation: a February 2026 ruling from the U.S. District Court for the Eastern District of Michigan, and a March 2026 ruling from the U.S. District Court for the District of Colorado, Morgan v. V2X, Inc. In citing these opinions, Judge Dorfman seemingly endorsed the rationale of those courts that the work product protection is waived only when materials are disclosed to an adversary, or in a manner substantially likely to reach one. The ruling thus suggests that the use of an AI tool did not fall within those parameters for waiver.

Notably, the judge “disagree[d]” with United States v. Heppner, a landmark February 2026 case from the U.S. District Court for the Southern District of New York. That case held that exchanges between a criminal defendant and publicly available generative AI platforms are not protected by the attorney-client privilege or the work-product doctrine. In that case, the federal court found that inputting sensitive information into a third-party consumer AI platform constitutes a voluntary disclosure outside the attorney-client relationship.

But Judge Dorfman stated that “the Texas rules set forth a different standard for protectable attorney work product and plainly appear on their face to extend that protection to” the plaintiff’s conversations with an AI tool.

Still, ultimately, the judge said the privilege only extends so far and ordered the plaintiff to hand over “all discovery materials or products that it has shared with” the AI tool, including materials produced under a protective order. The judge further recommended that the parties confer on amendments to a protective order “that would make unquestionably clear whether, how, and to what extent, if so, Confidential Information may be shared with any AI tool or other Large Language Model system.”

Key Takeaways

The Texas Business Court’s ruling comes as courts are grappling with, and reaching divergent conclusions about, how to handle privilege and parties’ exchanges with popular, publicly available generative AI platforms. Such issues are likely to grow as the use of generative AI tools becomes more ubiquitous.

Judge Dorfman’s minute entry, while not a full opinion on the merits, suggests that Texas courts may be more flexible in interpreting work-product protection for AI conversations, even when involving non-lawyer parties. Specifically, Judge Dorfman noted that the Texas rules of civil procedure protect parties’ materials and impressions developed in anticipation of litigation. The judge further suggested that discussions with a publicly available AI tool do not necessarily waive the privilege, as those conversations may not be substantially likely to be found by adverse parties.

Employers may want to review the extent to which their employees are using AI tools to make employment-related decisions and to make other decisions in response to employee complaints or the threat of litigation. Such materials may or may not be covered by the attorney-client privilege or the work-product doctrine, and disclosure of confidential information by non-lawyers to an AI tool may constitute a waiver of the privilege.

The case law to date suggests that individuals increase the risk of being found to have waived privilege when they use public AI tools, as opposed to enterprise-based AI tools. Additionally, beyond concerns of waiver, employers may need to evaluate whether generative AI tools constitute preservable electronically stored information (ESI) when issuing litigation holds. This is a quickly evolving issue on emerging technologies so employers may want to stay abreast of new developments.

Ogletree Deakins’ Technology Practice Group, Whistleblower and Ethics Practice Group, and Workplace Investigations and Organizational Assessments Practice Group will continue to monitor developments and will provide updates on the Employment Law, Ethics/Whistleblower, Technology, and Workplace Investigations and Organizational Assessments as additional information becomes available.

Follow and Subscribe
LinkedIn | Instagram | Webinars | Podcasts