Read the filing for Stansfield v. IBM. The order includes an explicit AI restriction against uploading to publicly available AI tools.

The filing

Stansfield v. IBM, U.S. District Court, District of Nevada, stipulated protective order entered April 16, 2026. 

What happened

A Nevada federal court approved a stipulated protective order in an employment dispute between Darin Stansfield and IBM. The order includes an explicit AI restriction: parties receiving confidential discovery materials may not upload them to publicly available generative AI tools such as ChatGPT unless those platforms can be configured to prevent the producing party’s data from being used for model training. Any AI-assisted litigation support must use secure, private tools only. 

Why it matters

Courts are beginning to codify AI data restrictions directly into discovery, creating real demand for enterprise AI tools that can prove confidential data never trains a public model. 

JetStream’s take

When a protective order names generative AI directly, AI governance stops being a policy question and becomes a discoverable, court-enforceable obligation. 

Source

Stansfield v. IBM – Protective Order (D. Nev.) by chelsea.stark.ctr