The new investigation will look into whether the company properly assessed and mitigated the risks of integrating Grok, particularly those of “manipulated sexually explicit images” including some that “may amount to child sexual abuse material,” the Commission said.
But the investigation “is much broader” than these images, a senior Commission official said during a briefing.
The chatbot may have generated as many as 3 million non-consensual sexual images and 20,000 child sexual abuse images in the 11 days before it made changes to stop the spread of such photos, an estimate by civil society found.
On top of the new investigation, the Commission will expand a 2023 probe to look into the impact of X’s decision, announced last week, to switch the algorithm for its social media platform to a Grok-based system.
The Commission said Monday it could take interim steps — for example, order X to change its algorithms or shut down the chatbot — “in the absence of meaningful adjustments to the X service,” something the EU has so far shied away from doing for Musk’s platform.
The threshold for such measures is “really high,” a second senior Commission official said.