LinkedIn Faces Lawsuit Over Data Misuse for AI

Microsoft ’s LinkedIn faces legal heat as Premium subscribers accuse the platform of sharing private messages with third parties to train generative AI models without their consent. A class action filed Tuesday in San Jose federal court alleges LinkedIn introduced a covert privacy setting last August, allowing users to opt out of data sharing—but only after the fact.

The complaint highlights a September 18 privacy policy update, where LinkedIn admitted user data could fuel AI training. In a linked FAQ, the company clarified that opting out wouldn’t undo prior data use. Plaintiffs claim this retroactive loophole demonstrates LinkedIn’s awareness of breaching privacy promises while trying to dodge public backlash.

Premium users allege their InMail communications were exposed to third-party entities before September 18, violating contracts and California’s unfair competition laws. They seek unspecified damages, including $1,000 per person for breaches of the federal Stored Communications Act.

LinkedIn dismissed the allegations as meritless, while plaintiffs’ legal counsel declined immediate comment. Notably, the lawsuit coincided with a major AI infrastructure announcement involving Microsoft-backed OpenAI, Oracle, and SoftBank Group Corp., underscoring the growing stakes in the AI arms race.