LinkedIn may have trained AI models with user data without updating its terms of service.
LinkedIn users in the U.S. (but not the EU, EEA or Switzerland, presumably due to data privacy regulations in those regions) have an opt-out toggle switch in their settings that reveals that LinkedIn scrapes their personal data to train “content creation AI models.” The toggle switch is not new; however, as first reported by 404 Media, LinkedIn did not initially update its privacy policy to reflect the data's use.
The terms of service have now been updated, but typically they are updated before any big changes like this one where they use user data for new purposes. The idea is to give users the option to change their account or leave the platform if they don't like the changes. That doesn't seem to be the case this time.
So what models does LinkedIn train? The company said in a Q&A that it trains its own models, including models for writing suggestions and post recommendations. But LinkedIn said generative AI models on its platform may also be trained by “separate providers,” such as its parent company, Microsoft.
“Like most features on LinkedIn, when you use our platform, we collect and use (or process) data about your use of the platform, including personal data,” the Q&A reads. “This includes your use of generative AI (the AI models used to create content) and other AI features, your posts and articles, how often you use LinkedIn, your language preferences, and any feedback you provide to our team. We use this data in accordance with our Privacy Policy to improve and develop LinkedIn services.”
LinkedIn previously told TechCrunch that it “uses privacy-enhancing techniques, including redaction or deletion of information, to limit the personal information included in the datasets used for generative AI training.”
To opt out of LinkedIn data scraping, go to the LinkedIn settings menu on your desktop.[データ プライバシー]Go to the section[Generative AI の改善のためのデータ]Click[コンテンツ作成 AI モデルのトレーニングにデータを使用する]Toggle the option off. You can also try a more comprehensive opt-out using this form, but LinkedIn says that opting out won't affect any training you've already taken.
Non-profit organization Open Rights Group (ORG) has called on the Information Commissioner's Office (ICO), the UK's independent regulator for data protection rights, to investigate LinkedIn and other social networks that train with user data by default. Earlier this week, Meta announced it would resume plans to collect user data for AI training after working with the ICO to streamline the opt-out process.
“LinkedIn is the latest social media company found to be processing our data without asking for our consent,” Mariano Delli Santi, ORG's legal and policy director, said in a statement. “Opt-out models have once again proven wholly inadequate to protect our rights. Ordinary people cannot be expected to monitor and track every online company that decides to use our data to train AI. Opt-in consent is not only a legal requirement, but also a common-sense requirement.”
The Irish Data Protection Commission (DPC), the supervisory authority that oversees compliance with the GDPR, the EU's comprehensive privacy framework, told TechCrunch that it was notified by LinkedIn last week that a clarification of the company's global privacy policy would be published today.
“LinkedIn has informed us that this policy includes an opt-out setting for members who do not want their data used to train content generation AI models,” a DPC spokesperson said. “This opt-out is not currently available to EU/EEA members as LinkedIn does not currently use EU/EEA members' data to train or fine-tune these models.”
TechCrunch has reached out to LinkedIn for comment and will update this article if we hear back.
The demand for more data to train generative AI models is leading to an increasing number of platforms repurposing vast amounts of user-generated content, and some even monetizing this content: Automattic, owner of Tumblr, Photobucket, Reddit, and Stack Overflow are among the networks that license their data to AI model developers.
Not all services make it easy to opt out: When Stack Overflow announced it would start licensing content, several users deleted their posts in protest, only for their posts to be reinstated and their accounts suspended.