X, previously known as Twitter, has updated its Privacy Policy to allow third-party companies to use X data for AI training, expanding beyond its own use by xAI’s Grok chatbot. Users can opt out, but the current settings don’t clearly specify where to manage this choice. This change becomes effective on November 15, suggesting additional options for opting out may be added.
Under the revised Privacy Policy, data retention is also modified. Instead of the previous limit of keeping “personally identifiable” data for up to 18 months, X now states that various types of information will be held based on service requirements, legal obligations, and security needs. The company notes that while content posted on X is kept for the account’s duration, public data, even after deletion, may remain in search engines or be retained by other entities, including those using it for AI purposes.
X has also added a “Liquidated Damages” clause to its Terms of Service, which enforces fees for content scraping. Any entity accessing more than a million posts in a 24-hour period may incur a $15,000 charge per million posts. This restriction aims to address scraping concerns as X seeks new revenue sources, following significant advertiser exits and lackluster interest in its paid subscription service. Like Reddit and some media organizations, X is exploring data licensing as an income stream.
