Microsoft’s LinkedIn is facing a lawsuit from Premium customers who claim their private messages were shared with third parties without consent to train generative AI models. 

According to Reuters, the proposed class action, filed on Tuesday night, accuses LinkedIn of secretly introducing a privacy setting in August 2024 that allowed users to control the sharing of their personal data.

The lawsuit alleges that LinkedIn then updated its privacy policy on September 18, 2024, to permit data use for training AI models. A “Frequently Asked Questions” link explained that opting out wouldn’t prevent data already used for training purposes. The plaintiffs argue this change was made discreetly, suggesting LinkedIn was aware it was violating users’ privacy while attempting to minimize public scrutiny.

Filed in a San Jose, Calif., federal court, the lawsuit represents LinkedIn Premium customers who sent or received InMail messages before September 18. The suit seeks damages for breach of contract and violations of California’s unfair competition law, as well as $1,000 per person for breaches of the federal Stored Communications Act.

Microsoft has not yet commented on the lawsuit. This legal action follows shortly after an announcement about a $500 billion AI infrastructure project involving Microsoft, OpenAI, Oracle, and SoftBank.