xAI and X respond to Grok criticism: Measures against lewd AI images

Elon Musk's AI development company xAI and the social network X are responding to persistent criticism of lewd AI images that could be created with the chatbot Grok using images of arbitrary individuals. In a statement, it is stated that a technological block will immediately prevent images of real people from being edited with Grok in such a way that they appear in bikinis or lingerie. Furthermore, image generation will only be available to paying users.

So far, however, this block does not seem fully effective. Even after the announcement, the chatbot responded to corresponding user requests, for example, one day later in the case of British Prime Minister Keir Starmer, who is depicted in a bikini by Grok at the request of a user.

Musk's Reaction

The fact that X is reacting at all is remarkable. Elon Musk had emphasized in an X post on Wednesday that the chatbot does not generate such depictions on its own, but merely responds to user requests. He also denied any knowledge that Grok was used to create nude images of minors using AI – one of the central accusations in the ongoing debate. However, Musk also stated in his X post that Grok generally respects the laws of the respective countries and will react with bug fixes to “attacks on Grok prompts.”

Musk himself said nothing about Grok creating lewd depictions of adults. In the X statement, however, a zero-tolerance strategy towards non-consensual nudity and unwanted sexual content is now also mentioned. Measures are being taken to remove content that violates these rules, and action is being taken against the accounts from which it originates, it states. In cases of sexual exploitation of children, law enforcement agencies would even be involved if necessary. In addition to the technical blocking of the chatbot function, geo-blocking has also been announced to exclude countries where such deepfakes are fundamentally prohibited.

Countries Took Action

Recently, more and more countries worldwide have taken or announced measures. Malaysia temporarily blocked Grok and sent formal complaints to X, California launched an investigation, and the EU Commission stated that it would apply the full toolkit of the Digital Services Act (DSA) if no measures were taken. The EU recently imposed a 120-million-euro fine on X for transparency deficiencies.

(mki)