LinkedIn has quietly introduced a policy update that allows the company to use user data dating back to its launch in 2003 to train generative AI models.
Unless users explicitly opt out before November 3, 2025, their profile information and public content will be included by default.
The data will be shared with Microsoft, LinkedIn's parent company, and its affiliates, under a legal justification called legitimate interest. No permission will be asked; opting out requires manual action through privacy settings. The change affects users in Europe, the UK, Canada, and Hong Kong and is part of broader privacy policy updates impacting over a billion global users.
LinkedIn's AI training data includes:
- Profile information: name, photo, work history, education, skills, location
- Public content: posts, articles, comments, polls
- Job-related data: resumes, application responses, and endorsements
- Feedback: ratings and reactions
- Private messages: explicitly excluded
Data from users under 18 and those with educational accounts is reportedly exempt. However, all other users are automatically opted in unless action is taken.
The Dutch data protection authority has issued a public warning, urging users to opt out before the deadline. AP vice-chair Monique Verdier emphasized the irreversibility of AI training, noting that once personal data is embedded into a model, it cannot be easily removed, and future use becomes difficult to trace or control.
The AP also highlighted the disproportionate risk when sensitive personal details such as health status, political views, or religious beliefs are involved, especially since this data may have been posted years ago without any expectation it would later be used for AI development.
While enforcement of EU privacy law (GDPR) falls under Ireland's Data Protection Commission (DPC), due to LinkedIn's European headquarters being based there, the AP has already received complaints and is working with the DPC and other European regulators to assess the legality of the company's plans.
Although the generative AI training clause doesn't currently apply to users in the United States, a parallel update to LinkedIn's terms will allow more of their data to be shared with Microsoft for personalized advertising starting November 3. This includes profile details, feed activity, and ad interaction data, depending on ad privacy settings.
How to opt out
To prevent your data from being used in LinkedIn's AI training, follow the steps below.
- Go to: Settings & Privacy → Data Privacy → Data for Generative AI Improvement
- Toggle off the setting
- (Optional) Submit a Data Processing Objection Form to formally object. This is also available to non-members if their data was posted by others.
Note that opting out only prevents future data from being used. Anything already processed will remain part of the training dataset.
Leave a Reply