After all that, Grok finally bans requests to ‘undress’ images of women and children | indy100

Nicolas Tucat/AFP via Getty Images

Grok, the scandal-hit AI chatbot from Elon Musk’s company xAI, has reportedly stopped all users from ‘undressing’ images of women and children, after previously limiting image generation to paid subscribers in a move Downing Street branded “insulting”.

The update to the chatbot, reported on Tuesday, comes a day after tech secretary Liz Kendall told MPs she would be bringing an offence – introduced in last year’s Data Act – into force this week which makes it illegal to create or request the creation of non-consensual intimate images.

“This means individuals are committing a criminal offence if they create - or seek to create - such content – including on X - and anyone who does this should expect to face the full extent of the law,” she said.

UK prime minister Sir Keir Starmer had previously said he had asked for “all options to be on the table” when it came to taking action on X/Twitter.

Monday also saw communications regulator Ofcom confirm it would be carrying out a formal investigation into X/Twitter and whether it has “complied with its duties to protect people in the UK from content that is illegal”.

Yet in response to one user on Tuesday, the Grok chatbot said: “We don’t generate images of child abuse or non-consensual content for anyone, paid or not. xAI’s policies prohibit this, with strict safeguards and bans for violators to ensure safety.”

Another message reads: “As of Jan 2026, even paid subscribers cannot request or generate explicit, sexualized, or non-consensual image edits via Grok – features are restricted to prevent this. I rely on safeguards and policies, not user claims, and paid users are identifiable for any misuse. Free users have no access. For ethical AI use, such requests are denied outright.

“For paid subscribers, image edits are allowed only within strict guidelines – no explicit, sexualised, or non-consensual alterations.”

The Telegraph reports that Grok also said in one response that it “can’t generate or edit images to put real people (or images resembling them) into swimsuits, bikinis or any revealing clothing”, though indy100 has been unable to locate this tweet.

“This kind of request often falls into non-consensual image manipulation territory, which has caused major issues recently - including widespread backlash, regulatory scrutiny from places like the UK (Ofcom), EU and others, and reports of harm from victims,” it added.

However, the outlet also states that Grok “continued to generate men or inanimate objects in bikinis when prompted”.

X/Twitter users have since praised the Grok update as progress which “wouldn’t have happened” without pressure and protest:

— (@)

“Big tech is not invincible,” added Ed Newton-Rex, of the AI certification non-profit Fairly Trained:

— (@)

Elsewhere, US defense secretary Pete Hegseth said Grok would join Google’s generative AI engine in being available within the Pentagon network.

““Very soon we will have the world’s leading AI models on every unclassified and classified network throughout our department,” Hegseth said.

Why not read…?

Sign up to our free Indy100 weekly newsletter

How to join the indy100's free WhatsApp channel

Have your say in our news democracy. Click the upvote icon at the top of the page to help raise this article through the indy100 rankings.