X to stop training Grok on user posts in EU
X will stop training its artificial intelligence (AI) chatbot Grok, using the public posts of users in the European Union and European Economic Area. This decision comes after facing pressure from a regional regulator. The company had temporarily halted this practice last month when Ireland's Data Protection Commission (DPC) initiated High Court proceedings against it. Now, X has made a permanent commitment, leading to the DPC dropping its legal action.
DPC's concerns over X's data protection practices
The DPC, acting as the primary EU regulator for X, expressed concerns that the company might have been infringing on data protection regulations and user rights. Since May, X had given users an option to opt out of their public posts being used for training Grok. This implied that this setting was automatically enabled for public accounts. However, under the EU's General Data Protection Regulation (GDPR), companies are generally required to get clear consent from users before processing their data.
European Data Protection Board's opinion sought
The DPC has also requested European Data Protection Board's input on key issues related to processing data for developing and training AI models. This includes how personal data is handled for such purposes. "The DPC hopes that the resulting opinion will enable proactive, effective and consistent Europe-wide regulation of this area more broadly," stated DPC commissioner Dale Sunderland in a statement. He added it would also assist in addressing complaints lodged with or forwarded to the DPC about these practices.
