LinkedIn has quietly opted users into using their data to train generative AI models without asking for consent upfront. A new privacy setting allows users to opt out, but any data used for training before the update is unaffected. Meaning, when LinkedIn introduced the toggle to opt out of AI model training, it clarified that any data already used for training purposes before the toggle was made available will not be affected by opting out. In other words, if your data was used in the past to train AI models, that data will remain part of the training sets, and there's no way to undo its use. The opt-out only stops LinkedIn from using your data for training models in the future, but all historical data prior to the update remains in use.
To stop LinkedIn from using your data for AI, users must navigate to the "Data for Generative AI Improvement" toggle in account settings and switch it off. Additionally, users must complete a separate form to prevent data from being used for other machine learning purposes like personalization and moderation. Data from users in the EU, EEA, or Switzerland is excluded from this process.
To turn off Data for Generative AI Improvments go here:
https://www.linkedin.com/mypreferences/d/settings/data-for-ai-improvement
LinkedIn Data Processing Objection Form: https://www.linkedin.com/help/linkedin/ask/TS-DPRO
Discussion (2)
The fact that the opt-out doesn’t apply retroactively is infuriating.
The classic "too little, too late" approach. Once your data is in, it’s in forever. How is this even legal in some regions? Or does GDPR only cover EU citizens? Feels like we’re being treated as data points instead of actual humans.
Do we even know what Facebook or Google are doing with all the data they’ve collected? I bet we’ll see more of this in the next couple of years. Get ready for every social media platform to quietly pull this same stunt.