Active

CNIL v. Meta (LLaMA training)

CNIL enforcement action challenging the lawful basis for training LLaMA models on EU resident data. The case tests whether legitimate-interest is an adequate GDPR basis for AI training at scale.

Docket snapshot

Full analysis in development. This case is tracked in our database with the metadata opposite, but a full editorial analysis has not yet been published. We prioritize deep-dive pages for cases with significant rulings or precedent value. If you have source documents or information that would accelerate publication, submit a tip.

Parties & claims

I · Overview

Commission nationale de l'informatique et des libertés v. Meta Platforms, Inc.

CNIL enforcement action challenging the lawful basis for training LLaMA models on EU resident data. The case tests whether legitimate-interest is an adequate GDPR basis for AI training at scale.

GDPRTraining DataLawful Basis

Why we’re tracking it

II · Significance

Privacy and GDPR enforcement cases determine whether large-scale AI training can rely on legitimate-interest theories or requires consent. The practical effect is global, not just European.

What this means for you

Status: Active. Last action: Feb 10, 2026. We update this page when the docket moves. For the full editorial analysis format, see our coverage of the Getty v. Stability AI, Andersen v. Stability AI, and NYT v. OpenAI cases.

Back to case tracker