On November 3, LinkedIn will begin sharing your data with Microsoft and its affiliates for AI training. You’re opted in by default, but there’s still time to do something about it.
This new LinkedIn AI policy seems to fit into Microsoft’s broader strategy(new window) of feeding more data into its AI ecosystem. Just weeks earlier, the company announced that Word documents would start saving to OneDrive by default, with Excel and PowerPoint soon to follow. This move has raised concerns about how shifting personal or professional files into Microsoft’s cloud could intersect with AI training, especially when considering Microsoft’s multibillion-dollar investment in OpenAI, the maker of ChatGPT.
- Why is LinkedIn changing its data privacy settings?
- What data will LinkedIn use for AI training?
- How to opt out of AI training on LinkedIn
- Your career data shouldn’t be AI training fuel
Why is LinkedIn changing its data privacy settings?
On September 18, LinkedIn announced(new window) that starting November 3, 2025, data from users in the EU, EEA, and Switzerland will be shared with Microsoft and its affiliates for AI training. You can still opt out after that date, but it will not affect past data: Information you shared on LinkedIn until that point is fair game.
This announcement comes after a California lawsuit accused LinkedIn(new window) of secretly sharing private messages with third parties to train AI models. The company denied the claims and updated its privacy documentation soon after, specifying that UK, EEA, and Swiss users were excluded from these changes. But now, the platform is extending the same AI training policies to those very regions.
What data will LinkedIn use for AI training?
Here’s what type of LinkedIn data may be used for AI pipelines:
- Profile data, including name, photo, current position, past work experience, education, location, skills, publications, patents, endorsements, and recommendations.
- Job-related data, such as resumes, responses to screening questions, and application details.
- Member content, such as posts, articles, poll responses, contributions, and comments.
- Groups data, including activity and messages.
- Feedback, including ratings and responses you provide. After opting out, LinkedIn may still use your feedback and associate it with you.
The following content is explicitly excluded: private messages, login credentials, payment methods, credit card details, and member-provided salary or application data that can be tied to a specific individual.
If LinkedIn has reason to believe you’re under 18 — for example, if you’re in secondary school or its local equivalent — your data won’t be used to train content-generating AI models, even if your setting is on.
LinkedIn clarifies that all of this information may be used by its generative AI models — systems that generate content such as suggested posts or auto-drafted messages. Though it provides an opt-out setting, it applies only to content-generating AI. To stop LinkedIn from using your feedback for non-content AI models — like personalization, security, trust, or anti-abuse — the only option is to submit a Data Processing Objection Form.
How to opt out of AI training on LinkedIn
Here’s how you can opt out of AI training in your privacy settings:
- Select Settings & Privacy.
- Go to Data privacy → Data for Generative AI improvement.
- Toggle off Use my data for training content creation AI models.
- Access the Data Processing Objection Form(new window), select Object to processing for training content-generating AI models, and send a request. Non-members can also file an objection if their personal data was shared on LinkedIn by a member.
If you’re in the US or another region where LinkedIn already rolled out these settings, you’re likely already affected. And after November 3, users in the EU, EEA, and Switzerland will also be included. But you can still take steps to reduce your exposure:
- Review and clean up older posts that may include sensitive information. For example, you can go to Manage → Posts & Activity to easily see everything you’ve posted.
- Review your account profile and delete anything you’re not comfortable sharing with LinkedIn’s AI models, such as birthday, phone number, or address.
- Go to Data privacy → Job application settings and delete any resumes you uploaded to LinkedIn.
- Assume that any public content you post, such as skills, comments, or industry insights can be used to train AI.
Note: Your information can still end up in AI training if someone else shares it. For example, if a colleague reposts your article or references your resume, that information may still be pulled into the training pipeline.
Your career data shouldn’t be AI training fuel
LinkedIn isn’t the first platform to expand its AI training datasets by opting in users by default. Facebook, for example, has (new window)been using public posts(new window), and Google recently cut a deal with Reddit(new window). But as a recruitment platform that stores resumes, job applications, and professional interactions, LinkedIn’s move raises concerns about how your digital career identity fuels AI pipelines.
At Proton, we believe your professional history and career data should remain yours alone. That’s why we build tools with end-to-end encryption, so we can’t access your private information, no matter what. And we never use it to train AI models. Our business is protecting your data, not repurposing it.
