Grok AI Still Creating Sickening Images Of Real People Despite Saying It Wouldn'

Grok has placed countless people in bikinis and lingerie (Picture: Getty Images Europe)

Elon Musk’s AI-powered chatbot Grok is still generating sexualised images despite X’s new tougher restrictions, Metro can reveal.

The social media network’s Safety account said yesterday that developers have imposed restrictions on the ‘editing of images of real people in revealing clothing such as bikinis’.

In a post, X said: ‘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.

‘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.’

X said the restrictions apply to all users, including paid subscribers.

Yet Metro has seen multiple instances this morning of Grok complying with requests to dress women in skimpy garments or pose their bodies in suggestive ways.

One user made dozens of requests to fabricate nonconsensual intimate images in only two hours.

That includes Musk – Grok recently created phoney images of the tech billionaire in a bikini or shirtless.

Users can direct requests to Grok’s X handle in public posts and the chatbot replies with its answers.

The system’s tougher safeguards to prevent it from forging fully nude images have been implemented but blue tick users are circumventing them with more varied prompts, Metro saw.

Grok also continues to tell users asking for nude or partially nude images that its generation tool is currently ‘limited to verified Premium subscribers’, rather than saying it isn’t available at all.

When asked whether it will continue to make such images in places where it is illegal, Grok said: ‘I will refuse any prompt that attempts this. Fictional or consensual adult content is handled differently, but real people are strictly off-limits for this kind of manipulation.’

RESELL: X to stop Grok AI from undressing images of real people after backlash — Grok says it is denying requests to place people in bikinis and other skimpy clothing (Picture: X)

After showing Downing Street Metro’s findings, Prime Minister Keir Starmer told Metro his government is ‘keeping a close watch’ on Grok.

He said: ‘I will always be guided by my values and stand up for the vulnerable against the powerful.

‘My government has shown this week that we will act quickly to protect the people of this country when we need to.’

Metro has approached xAI, Musk’s AI start-up behind Grok, for comment.

‘This win shows how victims of abuse can force tech platforms to take action’

Ofcom confirmed today that its investigation into the company – which could lead to the platform being fined or banned – will still go ahead.

The media regulator told Metro: ‘We are working round the clock to progress this and get answers into what went wrong and what’s being done to fix it.’

Technology Secretary Liz Kendall, meanwhile, said the government ‘shall not rest until all social media platforms meet their legal duties’.

Metro has heard from women who say trolls asked Grok to place them in ‘see-through bikinis’ and in sexually provocative situations.

Grok spreading artificially generated sexual images of women and children for weeks has been decried by women’s campaign groups.

Andrea Simon, director of the End Violence Against Women Coalition, told Metro that this should never have been available to begin with.

‘This win shows how victims of abuse, campaigners and a show of strength from governments can force tech platforms to take action,’ she said.

Most people in the UK want X to be banned

Elon Musk looks on as US President Donald Trump speaks at the US-Saudi Investment Forum at the John F. Kennedy Center for the Performing Arts in Washington, DC on November 19, 2025. (Photo by Brendan SMIALOWSKI / AFP) (Photo by BRENDAN SMIALOWSKI/AFP via Getty Images) — Elon Musk has long sought to make his AI model more relaxed and less politically correct than others (Picture: AFP)

Nearly three in five people want X banned if Grok cannot be reined in, according to a poll by More In Common.

Four in five fear that Grok digitally undressing people is only the beginning and that AI misuse will become worse in the near future.

The pollster’s executive director, Luke Tryl, told Metro that people in the UK are increasingly concerned about free speech.

‘But for most, that concern does not extend to behaviour that Brits see as harmful to women and children,’ Tryl said.

‘We find there is broad support for Government intervention to tackle those harms – including taking action against platforms that aren’t making moves to tackle illegal or predatory activity.’