Elon Musk has backed down in the row over the creation of deepfake images using his AI tool Grok, restricting the chatbot from creating sexualised images in places where it is illegal.
The use of Grok – an AI tool embedded into the social media platform X (Twitter) – to create sexually explicit content, including child sexual abuse, has prompted a wave of concern among ministers, with Ofcom earlier this week announcing an investigation into claims the tool was being used to digitally undress people.
The UK also said it would bring into force a law making the creation of non-consensual sexual images illegal after the row, in which Mr Musk claimed the British government was “fascist” and trying to curb free speech.
In a post on X on Wednesday, the company said it had imposed restrictions on “editing of images of real people in revealing clothing such as bikinis”.
“This adds an extra layer of protection by helping to ensure that individuals who attempt to abuse the Grok account to violate the law or our policies can be held accountable,” the company said.
“We now geoblock the ability of all users to generate images of real people in bikinis, underwear, and similar attire via the Grok account and in Grok in X in those jurisdictions where it’s illegal.”
The climbdown is a rare victory for Sir Keir Starmer, who was piling pressure on the tech billionaire to take action over the row.
The restriction will apply to all users, including paid subscribers, while image editing and creation will be limited to premium users.
Mr Musk claimed that his chatbot would refuse to produce illegal content and appeared to blame “adversarial hacking” for its generation of sexualised images.
With Not Safe for Work (NSFW) settings switched on, he said in an X post on Wednesday that the chatbot is “supposed to allow upper body nudity of imaginary adult humans (not real ones) consistent with what can be seen in R-rated movies on Apple TV”.
“That is the de facto standard in America. This will vary in other regions according to the laws on a country by country basis,” he wrote on Wednesday night.
The announcement came after Sir Keir said he “won’t back down” in his fight with Mr Musk over the creation of deepfakes – but welcomed reports that X had taken action to comply with the law.
Speaking at Prime Minister’s Questions on Wednesday, Sir Keir repeated his condemnation of Grok as “disgusting” and “shameful”, and said Ofcom had ministers’ full backing to act if X did not.
He added: “I have been informed this morning that X is acting to ensure full compliance with UK law. If so, that is welcome, but we’re not going to back down. They must act.”
A statement from the prime minister later added the government was “keeping a close watch on the situation”.
The PM was responding to Emily Darlington, Labour MP for Milton Keynes Central, who said: “If reports are correct, Elon Musk has climbed down today under pressure from this government. And let’s be clear, stripping women naked without consent in real life or online is abuse. But we don’t know whether to trust what X is saying today, and this isn’t just happening on X.”
Ms Darlington then revealed to The Independent that in the minutes after her question to the prime minister, people had put in requests to Grok asking for AI-generated pictures of her in a bikini or stripped naked.
She said: “Fortunately, Grok does not seem to have responded to them, so hopefully a change has been made.”
Also on Wednesday, the commons foreign affairs committee called on X, along with other social media firms, to answer questions from MPs in parliament about the threat posed by foreign disinformation targeting the UK online.
Mr Musk said he was not aware of “any naked underage images generated by Grok” and claimed the chatbot would “refuse to produce anything illegal” as its “operating principle” is to “obey the laws of any given country or state”.
