Grok chatbot trains on X user data in ‘very likely’ breach of EU law

Elon Musk could have yet another problem with the European Union.

Musk’s X has enabled the Grok chatbot to be trained on data from any user. This feature is now on by default, which may breach EU rules.

Deep within the settings of X, users unearthed evidence of the data harvesting. Besides a check box that was already ticked, they spotted the following text: “Allow your posts as well as your interactions, inputs, and results with Grok to be used for training and fine-tuning.”

This approach could violate the EU’s General Data Protection Regulation (GDPR). The law restricts companies from expanding data collection without telling users and providing a clear opt-out.

According to privacy group NOYB, it’s “very likely” that the setting infringed the law.

“Twitter/X should have informed its users properly and asked for consent,” the Vienna-based non-profit told TNW. The group is now assessing the details of the data collection.

X and Grok are not alone

X’s approach is reminiscent of a recent move by Meta. In June, the company was slammed for plans to train AI models on personal data — without requesting consent from users.

NOYB filed complaints in 11 EU countries about the scheme. In response, Meta claimed it has a legitimate interest for using the data.

The tech giant argued that it has a legal basis for training AI on publicly available information.

X may make a similar claim. Whether the EU agrees or not, users can turn the setting off. But even that isn’t particularly straightforward.

On the X mobile app, there’s no clear way to disable the data collection. Mercifully, the company has permitted deactivation on the web version.

You can either use this direct link or open X and navigate through Settings > Privacy and Safety > Grok. Then un-check the box that sends your data to Elon’s hungry bot.