LinkedIn silently opts users into AI model training, raises privacy concerns

LinkedIn has come under fire for automatically opting its users into a program that uses their data for training generative AI models. This revelation, first reported by 404Media, sparked privacy concerns among users who were unaware of the change.

Silent Opt-In and Updated Privacy Policy

LinkedIn introduced a new privacy setting and opt-out form without explicitly notifying users. The company then updated its privacy policy to state that platform data is being used for AI model training. As noted by TechCrunch, the policy has since been further revised.

LinkedIn's updated privacy policy now includes the following statement

"We may use your personal data to improve, develop, and provide products and Services, develop and train artificial intelligence (AI) models, develop, provide, and personalize our Services, and gain insights with the help of AI, automated systems, and inferences, so that our Services can be more relevant and useful to you and others."

Opting Out of AI Training

Users can revoke their consent for AI training by navigating to the "Data privacy" tab in their account settings. Under the "Data for Generative AI Improvement" section, a toggle switch allows users to opt-out. However, LinkedIn clarifies that opting out only prevents future data usage and does not affect training that has already occurred.

LinkedIn's FAQ page for AI training states that the company uses "privacy enhancing technologies to redact or remove personal data" from its training datasets. It also claims that data from users in the EU, EEA, and Switzerland is not used for training these models.

Beyond generative AI models, LinkedIn employs other machine learning tools for personalization and content moderation. To opt out of data usage for training these non-generative models, users must fill out a separate "LinkedIn Data Processing Objection Form."

Growing Concerns Over Data Privacy

LinkedIn's practice of silently opting users into AI training follows a recent admission by Meta that it has been scraping publicly available user data for model training since 2007.