AI in litigation: Texas Business Court addresses work product protection and discovery disclosure requirements

The court held that many AI-related communications by a company principal could be withheld as work product, but it also required the plaintiff to identify the discovery materials and products that had been shared with the AI tool, including materials produced under the parties’ protective order. Tate Grp. Auto. LLC v. Legacy Auto. Cap., LLC, No. 25-BC11B-0020, Ct. Minute Entry Regarding ChatGPT Materials In Camera Review, at 1-4 (Tex. Bus. Ct. 11th Div. June 3, 2026). For Texas businesses, the decision is an important reminder that using AI does not necessarily eliminate discovery protections, but uploading case materials to an AI tool may create separate disclosure, confidentiality and protective order issues.

What the court decided

The dispute arose after the defendants sought production of AI-related conversations between a non-lawyer principal of the plaintiff and a ChatGPT tool that the plaintiff had withheld as attorney work product. The defendants argued that work product protection should not apply to a non-lawyer’s chats with an AI tool, or that any protection was waived by the plaintiff’s use of AI. The court disagreed with the waiver argument, relying on Texas Rule of Civil Procedure 192.5, which protects “material prepared or mental impressions developed in anticipation of litigation or for trial by or for a party.” 

The court also agreed with two recent federal civil decisions, Warner v. Gilbarco, Inc., 2026 WL 373043 (E.D. Mich. Feb. 10, 2026), and Morgan v. V2X, Inc., 2026 WL 864223 (Mar. 30, 2026), that using a generative AI tool does not automatically waive work product protection. Those courts reasoned that a work product waiver generally requires disclosure to an adversary or disclosure in a way that substantially increases the likelihood that an adversary will obtain the protected material.

Disclosure of materials shared with AI

The most important part of the order for businesses is the distinction the court drew between AI communications and the materials uploaded to the AI tool as part of those communications. Even though the court held that using an AI tool did not automatically waive work product protection, it ordered the plaintiff to disclose “all discovery materials or products” shared with the AI tool, by Bates number if applicable, including materials produced under the protective order. This part of the order is perhaps surprising, as selection of key documents generally is one of the hallmarks of an attorney’s mental impressions and strategies; the disclosure to opposing counsel of which documents the plaintiff’s principal decided to share with the AI tool could provide a useful roadmap to plaintiff’s legal strategies. 

The court, however, may have been focused on potential confidentiality breaches by the documents being uploaded to an AI tool, as it stated that it would address any potential protective order violations if they were later shown. 
The court further recommended that the parties amend their protective order to make clear whether, how and to what extent confidential information may be shared with an AI tool or other large language model system. 

For businesses, the distinction between AI chat logs and the documents uploaded via the AI chat logs matters. AI communication may reflect litigation strategy and remain protected, but a court may take a different view regarding any documents, data or discovery materials uploaded to the AI tool and raise separate questions about confidentiality, protective orders, discovery obligations and company information policies. 

What this means for Texas businesses

The Tate Group Automotive order provides an early indication of how Texas courts may approach AI in discovery. Companies may wish to consult with counsel before using AI tools in litigation so that work product, confidentiality and protective order issues can be addressed and mitigated before information is uploaded. 

Businesses may wish to maintain AI protocols that identify approved tools, limit what information may be uploaded, require records of AI use in litigation, update legal hold and discovery procedures where appropriate, evaluate vendors’ data handling practices and address AI use in protective orders when appropriate. The key distinction is that an AI conversation may remain protected even when the documents or data submitted to the tool raise separate disclosure, confidentiality or protective order concerns.

As courts continue to draw lines between protected AI communications and the potentially discoverable materials shared with AI tools, businesses that set clear internal rules for litigation-related AI use now will be far better positioned than those forced to reconstruct their AI practices and account for AI-related records for the first time in the middle of a discovery dispute. 

__________

If you have any questions about this Legal Briefing, please feel free to contact any of the attorneys listed or the Eversheds Sutherland attorney with whom you regularly work.