LinkedIn, the professional networking platform owned by Microsoft, has unveiled significant updates to its privacy policy and user agreements. These updates, effective from November 3, 2025, introduce changes that will allow the company to share user data with Microsoft and its affiliates for training generative AI models and delivering personalized advertisements. While LinkedIn promises to respect user preferences, the default settings opt users into these data-sharing agreements unless they manually disable the feature.
AI Training and Data Sharing Policies
Under the new policies, LinkedIn will collect user data and content in specific regions, including Canada, Hong Kong, the European Economic Area (EEA), the United Kingdom, and Switzerland, to train its generative AI models. According to LinkedIn, these AI features are designed to help recruiters and businesses engage with users "more easily." The data used for AI training includes public content shared on profiles and posts, but the company has clarified that private messages will not be included in the dataset.
These AI-related policy changes, however, will not apply to users in the United States or the rest of the world.
Personalized Advertising and Expanded Data Sharing
In addition to AI training, LinkedIn has also updated its policies to enable broader data sharing for personalized advertising. For users in the United States, Canada, Hong Kong, and other non-EEA regions, LinkedIn will share additional data - such as profile information, feed activity, and ad engagement details - with Microsoft and its subsidiaries. Similarly, LinkedIn will use data from these affiliates to serve users with targeted ads.
LinkedIn’s definition of "affiliates" includes companies related by ownership, such as LinkedIn Ireland, LinkedIn Corp., Microsoft, and Microsoft’s subsidiaries. However, these advertising-focused policy changes will not apply to users in the EEA, the UK, and Switzerland.
Growing Industry Trend for Data Use in AI
LinkedIn’s updates reflect a broader trend across the tech industry, with multiple companies revising their policies to leverage user data for AI training. Platforms like Spotify, Meta, SoundCloud, WeTransfer, and Anthropic have all recently introduced similar changes to their privacy policies. These companies now use user data for purposes such as training AI models for recommendations or improving platform features.
For example, Meta revised its privacy policies in the European Union to include public posts and comments as training data for AI models. However, this decision has faced legal challenges, including a cease-and-desist notice from the advocacy group None of Your Business (NOYB). Similarly, SoundCloud and WeTransfer faced backlash over their decisions to allow AI training on user data, prompting further clarifications or revisions.
Opting Out of Data Sharing
While LinkedIn emphasizes that it will "honour" users’ prior settings and provide "robust controls" for data usage, the default opt-in mechanism places the responsibility on users to actively opt out of these data-sharing practices. Users who do not wish to participate in either AI training or personalized advertising can disable the default opt-in by navigating to the "Advertising Data" section within their account settings.
LinkedIn’s move to rely on default opt-ins aligns with industry practices but raises concerns about user privacy and consent. Critics argue that companies should seek explicit affirmative consent through opt-in mechanisms, rather than requiring users to opt out after the fact.
Conclusion
As the November 3 policy update approaches, LinkedIn’s changes to its user agreements highlight the increasing value placed on user data for AI development and targeted advertising across the tech industry. While LinkedIn provides users with options to control their data, the default opt-in approach underscores the ongoing tension between data utilization and user consent. Whether this shift will prompt further scrutiny or acceptance remains to be seen.
