Last August, LinkedIn "ostensibly" offered users the ability to enable a new privacy feature dictating whether their personal data could be shared, but turned permissions on by default, alleges the proposed class action lawsuit, filed in a California federal court. The change was not communicated in the social media giant’s terms of service or privacy policy before it was made.
However, the company "buried a crucial disclosure" in a frequently asked questions (FAQ) feature hyperlinked within the privacy policy, the lawsuit alleges, stating that users’ messages could also be used to train AI models by an unnamed third party provider, potentially outside of Microsoft. That disclosure, allegedly made only in the FAQ, suggests that LinkedIn was "aware its previous terms did not authorize these practices and was attempting to avoid further scrutiny".
LinkedIn also disclosed in the FAQ that the AI model training that had already been powered by users’ messages could not be reversed, the lawsuit alleges.
More: https://therecord.media/linkedin-lawsuit-private-messages-ai-training
