Microsoft × LinkedIn
On September 18, 2025, LinkedIn announced that starting November 3, 2025, it would use data from its members to improve content-generating AI models for itself and its parent company, Microsoft. [26, 35] The policy change makes user data, such as profiles and posts, available for AI training by default, though users can opt out. [33, 36] As this is an internal data transfer between a subsidiary and its parent company, no financial transaction price was announced. [35]
The asset
Use of LinkedIn user data, including public profiles, posts, and other activity, for training generative AI models. [26, 34, 36]
The record
- Seller status
- Going concern
- Sector
- social network
- Modalities
- text, images
- Personal data
- Users are opted-in by default but are provided with a settings page to opt out of future data use for AI training; the change does not apply retroactively. [33, 34, 36] The policy applies to users globally, with specific notices for different regions like the US and EU. [35, 36]
What it means if you hold data
This event highlights a significant method of data acquisition for AI training: internal transfers within large tech conglomerates. It bypasses direct licensing costs and leverages the vast user-generated content of a subsidiary platform (LinkedIn) for the benefit of the parent company's (Microsoft) AI ambitions. [26, 34]