On March 30, 2026, a Colorado federal court held that a pro se litigant’s AI-generated litigation materials are protected work product, then wrote its own protective order provision that effectively bars consumer AI platforms from processing confidential discovery materials.

Britannica and Merriam-Webster sued OpenAI for copyright infringement and added a Lanham Act claim alleging that ChatGPT’s hallucinations, when attributed to their brands, constitute false designation of origin.

Gracenote sues OpenAI for copying not just its media metadata, but the proprietary relational framework that organizes it. The complaint deploys established compilation copyright doctrine in a new context: as evidence that OpenAI's models encoded the protectable structure of a curated database, not just the data values.

The ruling underscores that no matter how egregious an employee’s conduct, a trade secret claim under the Defend Trade Secrets Act (“DTSA”) requires allegations that the defendant acquired, induced, or used the stolen information, not just that it hired the people who took it.

A federal magistrate judge in Michigan ruled that a pro se plaintiff’s AI-generated litigation materials are protected by the work product doctrine, and that using ChatGPT does not waive that protection. The decision creates a direct split with the Southern District of New York’s ruling in United States v. Heppner on the same issue, issued on the very same day.

On February 10, 2026, Judge Jed Rakoff of the Southern District of New York ruled that documents a criminal defendant generated using Anthropic's Claude AI are neither protected by the attorney-client privilege nor the work product doctrine.