Governments grapple with the flood of non-consensual nudity on X

For the past 2 weeks, X has been flooded with AI-manipulated nude images, created by the Grok AI chatbot. An alarming scope of women person been affected by the non-consensual nudes, including salient models and actresses, arsenic good arsenic news figures, crime victims, and adjacent world leaders. 

A December 31st probe insubstantial from Copyleaks estimated astir 1 representation was being posted each minute, but aboriginal tests recovered far more.  A illustration gathered from January 5th to 6th recovered 6,700 per hr implicit the 24-hour period. 

But portion nationalist figures from astir the satellite person decried the prime to merchandise the exemplary without safeguards, determination are fewer wide mechanisms for regulators hoping to rein successful Elon Musk’s caller image-manipulating system. The effect has go a achy acquisition successful the limits of tech regularisation — and a forward-looking situation for regulators hoping to marque a mark.

Unsurprisingly, the astir assertive enactment has travel from the European Commission, which connected Thursday ordered xAI to clasp each documents related to its Grok chatbot. The determination doesn’t needfully mean the committee has opened up a caller investigation, but it’s a communal precursor to specified action. It’s peculiarly ominous fixed recent reporting from CNN that suggests Elon Musk whitethorn person personally intervened to forestall safeguards from being placed connected what images could beryllium generated by Grok.

It’s unclear whether X has made immoderate method changes to the Grok model, though the nationalist media tab for Grok’s X relationship has been removed. In a statement, the institution specifically denounced the usage of AI tools to nutrient kid intersexual imagery. “Anyone utilizing oregon prompting Grok to marque amerciable contented volition endure the aforesaid consequences arsenic if they upload amerciable content,” the X Safety relationship posted connected January 3rd, echoing a erstwhile tweet by Elon Musk.

In the meantime, regulators astir the satellite person issued stern warnings. The United Kingdom’s Ofcom issued a connection connected Monday, saying it was successful interaction with xAI and “will undertake a swift appraisal to find whether determination are imaginable compliance issues that warrant investigation.” In a vigor interrogation connected Thursday, UK Prime Minister Keir Starmer called the improvement “disgraceful” and “disgusting,” saying “Ofcom has our afloat enactment to instrumentality enactment successful narration to this.”

In a station connected LinkedIn, Australian eSafety commissioner Julie Inman-Grant said her bureau had received a doubling successful complaints related to Grok since precocious 2025. But Inman-Grant stopped abbreviated of taking enactment against xAI, saying only, “We volition usage the scope of regulatory tools astatine our disposal to analyse and instrumentality due action.”

By acold the largest marketplace to endanger enactment is India, wherever Grok was the taxable of a ceremonial ailment from a subordinate of Parliament. On January, India’s communications regulator MeitY ordered X to code the contented and taxable an “action-taken” study wrong 72 hours — a deadline that was subsequently extended by 48 hours. While a study was submitted to the regulator connected January 7th, it’s unclear whether MeitY volition beryllium satisfied with the response. If not, X could suffer its harmless harbor presumption successful India, a perchance superior regulation connected its quality to run wrong the country.