The warning came as X users continued to generate thousands of pictures of women and children undressed using Grok’s AI, including in sexualised poses and in bikinis.
The so-called deepfakes have included images of the Princess of Wales, Cabinet ministers, MPs and celebrities.
On Wednesday, a UK internet watchdog warned it had uncovered images created with Grok that had been shared on a dark web forum that would constitute illegal child sexual abuse material.
It is understood that No 10 has raised the images directly with X, formerly known as Twitter, which Musk acquired for US$44 billion ($76b) in 2022.
Starmer told Greatest Hits Radio: “X has got to get a grip of this, and 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.”
The row threatens to deepen a rift with the United States over free speech. The White House has accused Labour of backsliding over freedom of expression, and President Donald Trump has directly criticised the UK’s tech rules as “not a good thing”.
Musk – who is the world’s richest man – has criticised Britain’s Online Safety Act, claiming the law’s purpose is the “suppression of the people”.
Under the act, British officials have the power to bar access to social media sites that repeatedly fail to take down illegal images such as child abuse material or revenge porn.
Ofcom, the technology watchdog, can apply for a court order – called an access restriction order – requiring internet companies such as BT or smartphone app stores to block offending sites and apps from Britain. So far, it has never used the power.
The regulator would be required to follow a legal process before applying for a ban, which would include an investigation and a provisional ruling. If a company refuses to address Ofcom’s concerns, it can seek to block the company’s website from the UK.
Ofcom warned this week that it could launch an investigation into X over the images and said it had made “urgent contact” with the social media site.
Researchers found that X users have been generating thousands of AI images of women per hour, which typically involve commenting beneath a picture with phrases such as “put her in a bikini”.
X’s Grok chatbot will then respond, generating a sexualised picture of a real woman without her consent.
Separately on Thursday, CNN reported that Musk had ordered staff at xAI, his AI business, to loosen the guardrails on Grok. A source said he had told a meeting he was “unhappy about over-censoring”.
Three xAI safety team members left the business soon after.
Louise Haigh, a former Transport Secretary under Starmer, urged the Government and the Labour Party to delete its X accounts.
She said: “The enablement, if not encouragement, of child sexual abuse mean it is unconscionable to use the site for another minute.”
X’s images have also provoked anger in the US. Ted Cruz, a Republican senator, called the AI pictures “unacceptable” and a violation of a US law banning AI revenge porn.
Musk said this week: “Anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content.”
X was contacted for comment.
Sign up to Herald Premium Editor’s Picks, delivered straight to your inbox every Friday. Editor-in-Chief Murray Kirkness picks the week’s best features, interviews and investigations. Sign up for Herald Premium here.
