LinkedIn Will Now Use Your Posts to Train AI Models by Default

LinkedIn (owned by Microsoft) will automatically use members’ data to train its generative AI models for content creation—unless users actively opt out. The platform has introduced a new setting that lets professionals decide whether their posts, messages, and other shared content contribute to improving LinkedIn’s AI-powered writing and content tools.

The setting, called “Data for Generative AI Improvement,” is turned on by default but allows users to opt out at any time. If switched off, LinkedIn and its affiliates will no longer use member-provided data to train models that generate content, such as the platform’s AI-powered writing assistant. However, the change does not apply retroactively to past training.

Importantly, the setting only applies to AI systems designed for generating content. Models used for personalisation, security, or abuse prevention remain unaffected. Members who opt out can still use LinkedIn’s generative AI tools, but their data won’t be used to improve them.

Feedback submitted by users — such as thumbs-up or thumbs-down on AI outputs — may still be used to refine content-generation models. Additionally, LinkedIn will exclude members it suspects are under 18 from AI training, even if their settings appear enabled.

"If we have reason to believe that you could be under 18 - for example, if you are currently in secondary school or its local equivalent (e.g., gymnasium or high school) - your data will not be used to train content generating AI models, even if your setting appears turned on," it said.

In the European Economic Area, Switzerland, and the UK, LinkedIn will update its Regional Privacy Notice effective November 3, 2025, to provide more details on AI training and opt-out options.