Dutch privacy regulator sounds warning over LinkedIn AI plans

And it warned greater regulation could be coming after talks with other data protection bodies across Europe.

The Autoriteit Persoonsgegevens (AP), which is the data protection agency for the Netherlands, said it had extreme concerns over the social media giant’s plans to utilise user data for training by default, meaning users would have to choose to opt out of the process manually.

LinkedIn said it plans to start using public posts, comments and user profile data – including names, photos, roles and skills – for “generative AI improvement” from 3 November.

“By default, and quietly introduced, all LinkedIn user data is being used for AI training purposes, which essentially forms an opt-out mechanism,” warned Nienke Kingma, a technology law expert at Pinsent Masons.in Amsterdam.

“Users have to actively disable the setting to prevent their data being used for these purposes. This opt-out approach has particularly caught substantial criticism from the regulators.”

The AP’s vice chair, Monique Verdier, said the regulator saw major risks in the plans announced by LinkedIn.

“LinkedIn wants to use data dating back to 2003, while people shared that information at the time without foreseeing that they would be used for AI training,” she said.

“Once that data is in an AI model, you lose control: it cannot be easily extracted, and the consequences are not easy to foresee. Care is certainly required with special personal data - such as data about health, ethnicity, religion or political preference.”

“That is why we call on everyone: adjust your settings before 3 November if you do not want data to be used for AI training.”

The AP said it was also working with Ireland’s privacy regulator, where LinkedIn’s European headquarters are based, and other regulators across the continent after receiving complaints about the move.