EU and UK attack X for sexual deepfakes generated by Grok

Last week, the Labour government had issued a harsh warning to Musk, asking the social network's top management to act with the utmost urgency to stop Grok's use of 'repugnant' deepfake images.

The European Union had also moved in this direction, condemning the use of chatbots and calling for X's intervention.

The decision to limit image editing to subscribers only was not described as satisfactory by the European Commission. In fact, the demand to put an end to the spread of deepfakes altogether remained valid.

The European Union will not delegate child protection and consent issues to Silicon Valley. This was stated by European Commission spokeswoman Paula Pinho during the daily press briefing, echoing the words of President Ursula von der Leyen regarding the function of Grok, the artificial intelligence chatbot developed by xAi and offered on the X platform, which allows users to generate sexually motivated deepfakes, even of minors, without the permission of the subjects in question. "This is unthinkable behaviour and the harm caused by these deepfakes is real.

We will not delegate child protection and consent to Silicon Valley. It is up to the company to address this chilling situation, otherwise we will intervene. We will not specify now in what form, but it is a concrete call,' says Pinho.

Article 70 of the European Digital Services Act specifies that 'proportionate measures' can be 'taken' to protect users in exceptional circumstances and where there is a risk of serious harm to them, adds another spokesperson, Thomas Regnier, emphasising that the EU law on digital services 'has already proved effective: we have received responses from X and the owner has announced measures that we are evaluating. Elon Musk is free to run his company as he wants in the EU, but compliance with our legislation is mandatory,' he emphasises.

Responding to another question about the perplexity, often originating in the US, about the use of DSA as a censorship tool, Regnier added that 'drawing a parallel between freedom of speech and an artificial intelligence tool that generates child pornography is dangerous nonsense, especially coming from the owner of a technology company.

Frankly, it is unbelievable that we have to discuss this in 2026. We are often accused of over-regulation, but this kind of content is exactly why we have laws. We have been clear: either X solves the problem, or we will take action under the legislation,' he concludes.

After Indonesia, Malaysia also decided on 11 January to suspend access to Grok, the artificial intelligence assistant of the social network X, which generates fake pornographic content from real photos. In detail, the Malaysian Communications and Multimedia Commission (McMC) stated in a statement that it had 'ordered a temporary restriction of access to Grok artificial intelligence for users in Malaysia, effective immediately'.

In the statement, the authority referred to 'content involving women and minors,despite a prior regulatory agreement and formal notices' sent to Elon Musk's company X and the start-up xAI that developed Grok. The Malaysian regulator said it considered X's protective measures insufficient, adding that access could only be restored once the required changes had been verified.

According to the Mcmc, X 'failed to respond to the risks inherent in the design and operation of the Ai tool', relying 'mainly on user-initiated reporting mechanisms'.

The decision comes as states around the world are calling for action against the system: from India, where hundreds of accounts were deleted today, to the EU, where access to Grok's services has been restricted to subscribers only.