X Updates Privacy Policy: Allowing Third Parties to Train AI Models Using User Data

X is updating its privacy policy to allow for the sharing of user data with third-party "collaborators" for the purpose of training AI models. Effective November 15, 2024, this new policy may pave the way for arrangements similar to those seen on Reddit, where external companies pay to license data from the platform.

The updated policy introduces a section labeled “third-party collaborators.” Depending on user settings and data-sharing choices, X may share personal information with third parties. If users do not opt out, these third parties may utilize the information for their own purposes beyond what is outlined in X’s Privacy Policy, including training generative AI models.

While the policy mentions an opt-out option, it is unclear how users can exercise this right. Tech media outlet reports that the policy directs users to their settings menu, but no apparent control for opting out of data sharing currently exists. However, since the policy does not take effect until next month, there may still be opportunities for changes. X has not provided a comment regarding this matter.

If X begins to license user data to other companies, it could create a substantial new revenue stream for the social media platform, which has recently experienced a decline in interest from major advertisers.

In addition to the privacy changes, X is revising its terms of service to impose stricter penalties on entities caught "scraping" large volumes of tweets. The section titled “liquidated damages” indicates that anyone accessing more than one million posts in a day will face a penalty of $15,000.

Protecting user data and platform resources remains a priority for X. Users acknowledge that, to the extent permitted by laws, if they violate the terms or assist others in doing so, they will be jointly liable for liquidated damages of $15,000 for any requests, views, or access exceeding one million posts, including replies, videos, images, and other content, in any 24-hour period.

Elon Musk has been vocal about issues related to scraping. Last year, he temporarily blocked access to tweets for users not logged in, citing it as a measure against scrapers. He also moved X's API behind a paywall, significantly impacting researchers' ability to study platform dynamics. Furthermore, Musk has leveraged allegations of scraping to support lawsuits against organizations investigating hate speech and other prevalent issues on the platform.