LinkedIn, the professional networking platform owned by Microsoft, is set to begin using its members' personal and professional data to train artificial intelligence models. The company has implemented an automatic opt-in policy, meaning users must actively change their settings before November 3, 2025, to prevent their information from being used.
Key Takeaways
- LinkedIn will share user data with Microsoft and its affiliates to train generative AI models.
- This is an opt-out system; users are automatically included unless they change their privacy settings.
- The deadline for users in specific regions to opt out and prevent data use is November 3, 2025.
- The data includes profiles, job applications, posts, comments, and other activity on the platform.
Details of the New Data Sharing Policy
LinkedIn has updated its terms of service to allow the sharing of user data for the purpose of training AI. The company justifies this action under the principle of “legitimate interest,” which allows it to proceed without requesting explicit permission from each user. This approach places the responsibility on individuals to navigate their privacy settings and manually withdraw consent.
The move is connected to Microsoft's significant investments in OpenAI, the developer of ChatGPT. Large Language Models (LLMs) like those developed by OpenAI require vast amounts of data to improve their performance and generate more accurate responses. The professional data available on LinkedIn is considered highly valuable for this purpose.
By making participation the default setting, the company is likely to gather a much larger dataset than it would with an opt-in system, where users would need to agree to participate actively.
Why is LinkedIn Data Valuable for AI?
LinkedIn's platform contains structured data on careers, skills, and professional interactions. This includes detailed work histories, educational backgrounds, skill endorsements, and industry-specific discussions. This type of information is ideal for training AI models to understand professional contexts, generate business-related content, and improve recruitment tools.
How to Opt Out of AI Data Training
Users who do not wish for their data to be used for AI training must take specific steps. It is important to note that opting out will only prevent the use of your data in the future; it does not retract any information that may have already been processed.
Step-by-Step Opt-Out Guide
- Log in to your LinkedIn account and navigate to the Settings & Privacy section.
- Select the Data privacy tab.
- Find the section titled Data for Generative AI Improvement.
- Locate the toggle switch for "Use my data for training content creation AI models" and turn it off.
For a more formal measure, users can also submit a formal objection to data processing. This can be done by accessing LinkedIn's Data Processing Objection Form, selecting the option to object to processing for AI training, and submitting the request. This option is also available to non-members whose data may have been shared on the platform by a member.
Important: According to the announcement, even after opting out, data already used for training will not be removed from the AI models. Users are also advised to review and remove any old or sensitive information from their profiles, posts, or resumes to limit their data exposure.
What Specific Data Will Be Used?
The scope of data LinkedIn plans to share for AI training is extensive. It covers nearly all aspects of a user's activity and personal information provided on the platform.
- Profile Information: This includes your full name, profile photo, current and past job titles, work experience details, educational background, location, listed skills, publications, patents, and any endorsements or recommendations received.
- Job Application Data: Information from resumes, answers to job screening questions, and details from applications submitted through the platform are included.
- User-Generated Content: All content you create or interact with, such as posts, articles, comments, responses to polls, and other contributions.
- Feedback Data: Any ratings, survey responses, or other feedback you provide on the platform may also be used.
Regional Differences and Affected Users
The policy changes affect users differently depending on their location. The official statement from LinkedIn specifies that members in the European Union (EU), European Economic Area (EEA), Switzerland, Canada, and Hong Kong have until November 3, 2025, to opt out.
While some reports suggest UK users are also included, it is recommended that any user who sees the opt-out setting in their account and wishes to decline participation should do so, regardless of their location.
"Starting November 3, 2025, we will share additional data about members in your region with our Affiliate Microsoft so that the Microsoft family of companies can show you more personalized and relevant ads."
For users in the United States, the terms update focuses more on data sharing for advertising purposes. The policy states that LinkedIn profile data, feed activity, and ad engagement data will be shared with Microsoft to deliver more personalized ads. This does not include data that users have already restricted for ad purposes in their settings.
Given that approximately a quarter of LinkedIn's over 1 billion users are in the U.S., this still represents a significant pool of data. All users are encouraged to review their advertising and data privacy settings to ensure they align with their personal preferences.