X Agrees to Limit Use of EU User Data for AI Chatbot Training, Reports Reuters

DUBLIN (Reuters) – Social media platform X has agreed to refrain from using personal data collected from EU users for AI training until users have had the opportunity to withdraw their consent, as stated in an Irish court on Thursday.

Ireland’s Data Protection Commission, which oversees compliance for many major U.S. internet companies operating in the EU, has requested the court to impose restrictions on X regarding the processing of user data for the development and training of its AI systems.

Under the ownership of Elon Musk, X claims that it provides users with the option to opt-out of having their public posts utilized by its AI chatbot, Grok. This can be done by deselecting a specific box in their privacy settings.

However, Judge Leonie Reynolds noted that X commenced processing data from EU users for AI training on May 7, while the option to opt-out was not made available until July 16. Additionally, this feature was not initially accessible to all users.

A representative for the platform, formerly known as Twitter, stated that any data gathered from EU users between May 7 and August 1 would remain unused until the court resolves the ongoing proceedings regarding the Data Protection Commission’s order.

Lawyers representing X are expected to submit opposition documents against the suspension order by September 4, as conveyed in court.

In a recent post, the X Global Government Affairs account criticized the regulator’s request for an order, labeling it as unjustified and excessive, suggesting that it unfairly targets X.

The regulator’s scrutiny of X’s data practices comes in the wake of Meta Platforms’ decision in June to postpone the launch of its AI models in Europe after being advised by the Irish Data Protection Commission. Similarly, Google’s Gemini AI chatbot faced delays earlier this year following discussions with the Irish authority.