LinkedIn is now facing a lawsuit for using customer data and communications to train AI models, according to a new proposed class-action filing submitted in California this week.

A LinkedIn Premium subscriber is suing the Microsoft-owned networking platform because they believe LinkedIn has been "unlawfully disclosing" their private direct messages (DMs) to third parties for the purpose of training AI. The lawsuit accuses LinkedIn of violating the Stored Communications Act, Breach of Contract, and California's Unfair Competition Law.

Last year, LinkedIn quietly started training AI models on your LinkedIn data—and added a button in the settings menu that lets you opt out. But this setting is on by default, meaning some users may not have been aware that their data was being used automatically. Shortly after this announcement, UK regulators raised user data privacy concerns, and LinkedIn stopped training on UK-based users. LinkedIn said it also isn't using EU user data or Switzerland-based user data.

At the time, LinkedIn didn't specify exactly who it's sharing user data with to train said AI tools, instead stating that its "affiliates" would be given the data. When I previously pressed LinkedIn for clarification, a spokesperson told me that "affiliates" refer to any Microsoft-owned company (but not Microsoft-backed OpenAI). That said, however, Microsoft has acquired more than 270 companies since 1986, including five AI companies, so it remains unclear as to who exactly is using this data.

The lawsuit suggests that "another provider" is using the LinkedIn customer data for AI training. "Private discussions could surface in other Microsoft products, and customers’ data is now permanently embedded in AI systems without their consent, exposing them to future unauthorized use of their personal information," the complaint argues, adding: "LinkedIn has not offered to delete the data from the existing AI models or retrain them to eliminate their reliance on the disclosed information."

LinkedIn, however, denies the claims in the lawsuit, telling multiple news outlets: "these are false claims with no merit."

The filing is named De La Torre v LinkedIn Corp, US District Court, Northern District of California, No. 25-00709. The Plaintiff is seeking $1,000 in damages and potentially other relief as compensation.

If you want to opt out from LinkedIn using your data to train AI, you can go to Settings > Data Privacy > Data for Generative AI Improvement and turn the toggle off.