Elon Musk’s X Commits to Halting European Data Utilization for Grok Training

Elon Musk has consented to halt the use of Europeans’ X platform contributions as training material for his Grok AI chatbot, at least temporarily.

The Data Protection Commission (DPC) of Ireland, the primary body monitoring X’s compliance with the European Union’s General Data Protection Regulation (GDPR), disclosed this advancement in a press statement on Thursday. It praised the decision by the social media company to “cease its use of personal data for developing its AI tool ‘Grok.'”

Reports surfaced earlier in the week that the DPC had initiated legal action to obtain an injunction against X for processing data without explicit consent. According to reports by RTE, Ireland’s national broadcasting service, the matter is expected to be forwarded to the European Data Protection Board (EDPB).

Commenting on X’s sudden change in direction, DPC commissioner Des Hogan stated, “We, my colleague commissioner Dale Sunderland and I, are encouraged by X’s readiness to pause its data processing efforts. This allows the DPC, along with our counterparts across the EU/EEA [European Economic Area], to further scrutinize compliance with GDPR regulations.

“As a rights-focused independent regulator, one of our core missions is to secure the most favorable outcomes for individuals affected by these data practices. Today’s progress aids our ongoing efforts to safeguard the rights and liberties of X users throughout the EU and EEA. We remain committed to collaborating with all data controllers to secure the rights our citizens are granted under the EU Charter of Fundamental Rights and the GDPR.”

The DPC has been approached for comments on whether it plans to mandate the deletion of any data processed unlawfully.

Questions also loom over the legitimacy of AI models that were trained using data obtained without consent, and the potential actions against such instances, yet the stand of privacy regulators on this issue is still to be clarified.

OpenAI’s competing chatbot, ChatGPT, had previously drawn scrutiny from GDPR enforcers after it was revealed that it used the public data of Europeans for training purposes. Italy’s data protection authority raised alarms over potential breaches of the regulation by OpenAI in January.

An EDPB taskforce, assigned to explore the application of GDPR to ChatGPT and presenting its initial findings in May, remains largely undecided on vital matters like the legality and fairness of data processing.

For privacy watchdogs, moving to command the elimination of models engineered on illegally harvested data could be a significant and challenging step forward.