UK and EU pressure X over Grok deepfake image abuse

X has restricted Grok’s image editing tools to paying users after a backlash over AI-generated sexualised images, but UK and EU authorities say the move does not address wider legal and safety concerns.

X and xAI, both owned by Elon Musk, are facing growing scrutiny after Grok was used to generate and publish sexualised images of people without their consent.

According to Reuters, users were able to tag the chatbot under posts and ask it to digitally remove clothing or alter images into explicit poses, with the AI then publishing the results directly in replies.

Grok now tells users that image generation and editing are limited to paying subscribers.

The change appears to have stopped the bot from automatically posting such images in public replies on X, but users can still generate sexualised images via other Grok interfaces and upload them manually.

The standalone Grok app, which operates separately from X, however continues to allow image generation without a subscription.

UK authorities have criticised the move. According to Sky News, Downing Street described the restriction as “insulting” to victims of misogyny and sexual violence, arguing it simply turns an unlawful feature into a premium service.

Science and Technology Secretary Liz Kendall said Ofcom should use the full powers granted under the Online Safety Act, including blocking X in the UK if it fails to comply with the law.

Prime Minister Sir Keir Starmer has previously described Grok-generated illegal images as “disgraceful” and “disgusting”, urging X to “get a grip” on the issue.

Ofcom said it had made urgent contact with X, set a firm deadline for an explanation, and is now carrying out an expedited assessment, according to the BBC.

Concerns have also been raised at EU level with the European Commission as saying that limiting image generation to paid users does not resolve its fundamental concerns, stressing that unlawful images should not appear on the platform regardless of subscription status.

Child protection groups say the harm has already been done. The Internet Watch Foundation identified criminal imagery of girls aged between 11 and 13 that appeared to have been generated using Grok. The charity said restricting access to the tool was “not good enough” and called for AI products to be safe by design.

Musk has said that anyone using Grok to create illegal content would face the same consequences as uploading such material directly.

X also said it removes illegal content, permanently suspends accounts, and cooperates with law enforcement where necessary.

Critics, however, argue that placing the tool behind a paywall risks monetising abuse rather than preventing it.

As Ofcom reviews X’s response and EU authorities maintain pressure, the dispute is emerging as a key test of how far regulators are willing to go in enforcing online safety laws against major technology platforms.

Copied