"Disgusting": Criticism of X grows over Grok's sexualized deepfakes

The European Union and authorities in an increasing number of states are investigating the microblogging service X and the Grok AI image generator offered there because it allows for the direct and public creation of sexualized deepfakes of people, including children. Such results from the so-called “Spicy Mode” are not “spicy” but illegal, explained an EU Commission spokesperson: “This is outrageous, this is disgusting.” The Commission is already in contact with Elon Musk's microblogging service due to previous scandals involving Grok and is taking the new matter very seriously. Authorities in Great Britain, India, and Malaysia have also launched investigations, and France has described the generated images as “clearly illegal.”

Outrage all over the world

The background to the recent outrage surrounding Grok are users on X who have used the available AI account to digitally “undress” photos of women and even minors and generate sexualized versions of them. The images appear publicly on the microblogging service. X has stated through the account itself that these were only “isolated cases” and that a causal “failure of security measures” has been fixed. Nevertheless, the AI continues to fulfill such user requests days after the public became aware of the behavior, and the AI's profile is still full of such images.

In Great Britain, the media regulator Ofcom has established “urgent contact” with Elon Musk's xAI, which is now responsible for X and Grok, reports the BBC. In India, an investigation was already initiated before the weekend, and the US company was given a deadline until yesterday, Monday, to respond, writes CNBC. In Malaysia, authorities have also become active and announced that they will summon representatives of X. According to the report, calls have already been made in the Brazilian parliament to block X nationwide until an investigation is concluded.

Elon Musk himself claimed over the weekend that anyone using the AI to create illegal content would face the same consequences as for uploading illegal content. However, at present, it does not appear that this will actually have consequences. When Grok began spreading antisemitism and Holocaust denial last year, the AI account was temporarily muted. At that time, some objectionable content was removed. X apparently does not consider this necessary this time, and according to CNBC, no harsh statements have yet come from the responsible US authorities like those from other states or the EU.

No new phenomenon

The fact that people are using the available and increasingly sophisticated AI image generators to digitally “undress” individuals – i.e., to create fake but realistic-looking sexualized images – is not a new phenomenon. For a long time, this happened largely out of the public eye, even though a lucrative business field has developed from it years ago. However, this was only possible because those responsible for the technology charged money for it. The fact that this is now happening in full public view on X, is completely free of charge, and appears to remain without consequences is a new quality. Affected women report a feeling of dehumanization when their photos are sexualized in this way without them being able to do anything about it.

(mho)