Grok turns off image generation for most users after it removed clothing from children

X has stopped its AI system for generating images for most of its users – after it was used to generate abusive pictures of women and children for days.

The Grok system has brought widespread outcry in recent days after it was used to repeatedly generate sexually explicit and violent images.

Users are able to ask the system to make a picture simply by sending it a tweet. That led to a trend in which users send the system an image of a clothed person – occasionally underage – and asked it to undress them, which it did.

It meant that more than half of the images being produced by the system were of people with their clothes digitally removed, according to recent analysis.

The Grok account is now however replying to requests for images with a message indicating that it is only available to some users. “Image generation and editing are currently limited to paying subscribers,” the message reads, and it gives users a link to sign up for the premium version of X, formerly known as Twitter.

The change comes amid criticism from governments and suggestions that the app could be banned in some locations entirely. Europe and the UK have both launched investigations into the platform and threatened to shut it down.

Keir Starmer said he backed Ofcom, which is looking into X and xAI – the firm founded by Elon Musk which created Grok – to take action against the site, and that he had asked for “all options to be on the table”.

“This is disgraceful. It’s disgusting. And it’s not to be tolerated,” the prime minister said.

“X has got to get a grip of this. Ofcom has our full support to take action in relation to this. This is wrong.

“It’s unlawful. We’re not going to tolerate it. I’ve asked for all options to be on the table.

“It’s disgusting. X need to get their act together and get this material down. We will take action on this because it’s simply not tolerable.”