AI | LinkedIn under scrutiny for using user data to train AI

Starting from 3 November 2025, LinkedIn is using its users’ data to train its artificial intelligence (AI) models, following updates of its terms of service and privacy policy. The news has been released by the platform on its blog and through e-mails and in-app notifications to its users in the past days.

The updates will apply to all users that did not exercise their opt-out right before 3 November 2025. For those who exercised the opt-out right afterwards, LinkedIn will stop processing their data for this purpose going forward; however, data already collected will continue to be used to train the AI models. Users have been informed that their data may be used to “gain insights with the help of AI, automated systems, and inferences”. LinkedIn has also released some FAQ that can be used to better understand the activity and its purpose, as well as guidance on the company ‘s approach to generative AI.

In response, the Italian Data Protection Authority (Italian DPA) released an informative notice clarifying, inter alia, how Italian users can exercise their right to opt-out.  The Italian DPA also announced coordinated action with other European data protection authorities – led by the Irish authority – to assess LinkedIn’s compliance with the GDPR, including the adequacy of its opt-out mechanism, the categories of personal data involved and the legitimacy of its reliance on “legitimate interest” as a legal basis. The initiative follows similar scrutiny faced by Meta earlier this year, as European authorities continue to question whether large platforms’ use of personal data for AI training complies with data protection principles.

Newsletter n. 112 – November 2025