Court Greenlights Generative AI for Legal Discovery: Three Key Rulings from Schulte v. LinkedIn

A June 30, 2026 ruling from the Northern District of California has given enterprises meaningful legal footing for deploying generative AI in discovery workflows. In Schulte v. LinkedIn Corp., the court reviewed LinkedIn's use of Relativity aiR, a generative AI-powered document review tool, and rejected all plaintiff challenges to its use.

The decision signals that courts are ready to treat generative AI review as an extension of established discovery practice rather than a novel legal category requiring new doctrine.

Three Key Takeaways

1. Generative AI review is treated as TAR, not a new category. The court applied existing Technology Assisted Review (TAR) principles directly to LinkedIn's AI-assisted workflow. Existing proportionality, reasonableness, and transparency standards apply. No wholesale rethinking of discovery rules is required.

2. Search-term pre-filtering before AI review is permissible. Plaintiffs challenged LinkedIn's use of search strings to reduce the document population before feeding it into the AI tool. The court rejected the challenge, consistent with existing TAR case law: search terms may be used to reduce review populations before AI-assisted review begins.

3. Discovery on discovery remains disfavored. Opposing parties seeking detailed AI performance metrics must demonstrate a specific production deficiency, not just speculate that the AI process may be inadequate. This gives producing parties meaningful protection when using well-designed AI workflows.

WilmerHale recommends that parties seeking access to AI metrics include particularized disclosure requirements in ESI protocols before discovery begins.

Read the full article on WilmerHale