Dutch privacy watchdog urges LinkedIn users to opt out of AI data training by Nov. 3

The Dutch Data Protection Authority is urging all LinkedIn users to review and adjust their privacy settings, because the social media platform will soon begin using personal data to train artificial intelligence (AI). The privacy authority said on Wednesday it has “major concerns” about the website’s plans, which requires users to explicitly refuse to be included in the data mining operation.

LinkedIn announced it will use various pieces of data for AI starting on November 3, including names, photographs, and work experience. Anyone who does not want this data to be used must opt ​​out in their settings before that date. Public posts, comments, and polls will also be included.

“We see significant risks in the announced plans,” says Dutch DPA Vice-Chair Monique Verdier. “LinkedIn wants to use data dating back to 2003, even though people shared that information at the time without foreseeing that it could be used for AI training,” she continued.

“Once that data is in an AI model, you lose control: it’s impossible to remove it, and the consequences are difficult to predict.”

The regulator believes care is especially needed with sensitive pieces of personal information, such as data on health, ethnicity, religion, or political affiliation. “Therefore, we urge everyone to adjust their settings before November 3rd if you do not want data to be used for AI training,” Verdier stated.

According to the Dutch regulator, it’s not yet certain that LinkedIn will be allowed to do what it plans to do. LinkedIn falls under the supervision of the Irish Data Protection Authority, because that is where the company has its European headquarters.

The Dutch authority, known locally as the Autoriteit Persoonsgegevens, said it has already received complaints about the social media network’s planned use of data. These will be addressed in cooperation with the Irish authorities.