LinkedIn sued for sharing private messages to train gen AI models

Professional social network LinkedIn has been sued by its premium customers for “breaching its contractual promises” and allegedly sharing their private messages with third parties to train generative AI models.

The lawsuit, filed in a California court, argues that the personal data of LinkedIn’s premium customers is protected by the LinkedIn Subscription Agreement (LSA), which promises “not to disclose premium customers’ confidential information to third parties.”

LinkedIn refuted the allegations in a statement that said, “these are false claims with no merit.”

In August 2024, LinkedIn quietly introduced a privacy setting with an opt-out option that allowed users to disable sharing of their personal data for training AI. A month later, LinkedIn updated its privacy policy, informing users that the platform is using their personal data to improve, develop, and train generative AI models. The lawsuit alleges that this was done to mitigate backlash and legal repercussions. 

The lawsuit further claims that the privacy setting was enabled by default allowing LinkedIn to harvest user posts and messages to train AI models without the knowledge or permission of users. LinkedIn users in the EU, UK, Canada, China, Switzerland, and Hong Kong were exempt from the data collection. 

The US still doesn’t have a federal data privacy law like the EU’s General Data Protection Regulation (GDPR) or India’s Digital Personal Data Protection Act (DPDPA) 2023 that gives users more control over their personal data and guarantees protection against unauthorized data collection practices. 

“Given its role as a professional social media network, these communications include incredibly sensitive and potentially life-altering information about employment, intellectual property, compensation, and other personal matters,” the lawsuit claims. 

Founded in 2002 and acquired by Microsoft in 2016, LinkedIn claims it has over a  billion members. According to Statista, the US is its biggest market with 230 million users followed by India with 130 million users. Premium subscribers are estimated to be around 39% of LinkedIn’s user base. 

The lawsuit alleges that the personal data shared by LinkedIn with third parties for AI training includes InMail messages. The lawsuit contends that LinkedIn has not publicly refuted the claim that premium InMail messages were included in this data sharing.

LinkedIn InMail is a premium messaging service that allows users to send messages to members who are not in their network connections. 

The growing interest in generative AI applications has fueled a demand for high-quality training data. LinkedIn-parent Microsoft, which is a major investor in OpenAI, has integrated OpenAI products across its services, including Azure Cloud, Bing Search, Windows 11, and Microsoft 365. 

Many companies at the forefront of generative AI development are facing legal challenges over their data collection practices. 

Perplexity AI was sued last October by The Wall Street Journal, New York Post, Financial News, and Barron’s for allegedly using their copyrighted news articles without permission to train its AI models. Eight US newspapers including the New York Times have also sued OpenAI and Microsoft for similar copyright infringement. The two firms are also facing similar lawsuits from writers who claim their work was used to train AI models without their consent. 

Image credit: Pexels