September 2026
Outline
A look at pertinent announcements, updates, and reports that affect discovery.
GenAI Prompts: A Signal Worth Watching
A Texas Court Shields Generative AI Review Prompts as Work Product, but Leaves No Written Opinion Behind | Law.com
For lawyers watching how courts will treat attorney-crafted generative AI instructions, the outcome is instructive. But the way it was delivered says just as much about where this area of law actually stands.
A District Court Judge in the West District of Texas ruled that prompts created by attorneys were work-product and likely privileged. This echoes recent decision showing a trend toward protected attorney-crafted prompts and instructions.
In case you missed it: United States v. Heppner revisits attorney-client privilege in the world of AI
Harvard Law Review
United States v. Heppner Harvard Law Review
Fitting new technology to old doctrine is a perennial challenge for courts. Today, that technology is generative artificial intelligence (AI), and those doctrines now...
The court ruled that information provided to the attorney which was first put into a public AI tool by the client without attorney consent was neither privileged nor work-product. The main drivers of the outcome was the combination of the absense of attorney direction and the use of a consumer AI product with terms of service that allowed for inputs to be used for training. The court held that using AI in that instance was akin to a third-party disclosure, which disposed of any claim of privilege.