Elon Musk’s X Allows AI Training on User Data with New Privacy Policy

On Wednesday, social network X (formerly Twitter) updated its Privacy Policy to reflect a significant change. The platform will now allow third-party collaborators to train their AI models on X data, unless users choose to opt out. This marks a shift from the company’s previous policy, which did not explicitly include third-party AI training.

Third Parties Gaining Access to X Data

Previously, Elon Musk’s AI project, xAI, trained its Grok AI chatbot using X user data, which triggered an investigation by the European Union’s privacy regulators. However, X had not updated its policy to state that third parties could access and utilize this data. Now, like Reddit and various media organizations, X is exploring licensing its data as a new revenue stream, signaling its intent to offer this information to AI companies.

What’s New in X’s Privacy Policy?

The changes appear in Section 3 of X's updated Privacy Policy under “Sharing Information.” A new paragraph explains how user data can be shared with third-party collaborators, and how users can opt out if they prefer not to participate.

The section states:

“Third-party collaborators. Depending on your settings, or if you decide to share your data, we may share or disclose your information with third parties. If you do not opt out, in some instances the recipients of the information may use it for their own independent purposes in addition to those stated in X’s Privacy Policy, including, for example, to train their artificial intelligence models, whether generative or otherwise.”

The policy directs users to their settings to opt out, though as of now, it’s unclear where this specific option is located. Currently, the “Privacy and safety” settings only provide options to turn off data-sharing with xAI’s Grok and certain business partners, but these partners are described as those helping to “run and improve its products,” rather than third-party AI providers.

When Will the New Policy Take Effect?

The updated privacy policy is set to go into effect on November 15. By then, users should expect an opt-out option to be available, though this remains speculative for now.

Changes to Data Retention Policies

In addition to changes surrounding AI training, X has removed previous assurances regarding how long it retains user data. The old policy stated that user profile information and content would be kept for the duration of the account and that personally identifiable data collected during service use would be retained for up to 18 months.

The updated policy instead specifies that X will retain different types of information for various periods based on their necessity for providing services, complying with legal obligations, and ensuring safety. For example, content like posts and interactions will now be kept for as long as the account remains active or until the content is removed.

AI and Public Data Retention

Interestingly, the updated policy also reminds users that public content can persist beyond X’s platform, particularly if it has been ingested by AI providers. X notes, “search engines and other third parties may retain copies of your posts longer, based upon their own privacy policies, even after they are deleted or expire on X.”

New Liquidated Damages Clause for Scraping

In another update, X has introduced a new “Liquidated Damages” section in its Terms of Service. Organizations scraping X content will now be liable for financial damages. Specifically, X states that viewing or accessing more than 1,000,000 posts in a 24-hour period will result in a charge of $15,000 USD per million posts.

X's Financial Challenges

The monetization of X’s data comes amid financial struggles, following advertiser boycotts and an underperforming subscription feature. The company is actively seeking alternative revenue streams to sustain its operations. As of now, X has not responded to requests for comment on these changes.