EU flags ‘appalling’ child-like deepfakes generated by X’s Grok AI

The alleged spread of explicit, child-like material on social media platform X has been condemned by the European Commission, which labeled the content “appalling” and “disgusting.”

Following weeks of complaints about a new feature on X’s integrated AI chatbot Grok that included depictions of children, European Union spokesman Thomas Regnier made the comments to reporters on Monday.

Regnier added that “Grok is now offering a “spicy mode” that features explicit sexual content with some output produced by child-like images. This is not spicy, he said. This is prohibited. This is terrible. This is repulsive.

Regnier claimed that such content has no place in Europe and that the European Commission is “very seriously looking” into it.

In addition, the public prosecutor’s office in Paris, France expanded its scope of the investigation into X to include allegations that Elon Musk’s xAI company, Grok, was used to create and spread child porn.

“Waiting to be armed.”

Users on Grok had the option to edit any image on the platform in the late December when a novel “edit image” feature was released. Some users requested that Grok remove women or children from photos in part or in whole.

Users frequently request that users alter their photos of women so they can be seen wearing revealing outfits, like translucent bikinis.

On Friday, Grok acknowledged there were “lapses in safeguards” and stated that it was “urgently fixing them.” It stated in a post that “Casim[Child Sexual Abuse Material] is prohibited and illegal.

However, according to AI safety experts, the platform ignored months of admonitions that such abuse was on the rise.

According to Tyler Johnston, executive director of AI watchdog group The Midas Project, “we warned in August that xAI’s image generation was essentially a nudification tool waiting to be weaponized.” That is essentially what is being played out, he said.

The EU has already criticized X, which Musk purchased in 2022, for its advertising and content-moderation policies.

For using fraudulent methods for user verification and in violation of the EU’s digital content regulations, Brussels fined the platform 120 million euros ($140 million) in December.

In a probe that began in December 2023, the EU’s Digital Services Act is still being investigated.