Ofcom begins investigation of specific picture era on Grok

On-line security regulator Ofcom has begun a proper investigation into X below the UK’s On-line Security Act, following what’s being considered misuse of the Grok AI chatbot.

The regulator mentioned it was responding to what it says is deeply regarding reviews of the chatbot account on X getting used to create and share undressed photos of individuals – which can quantity to intimate picture abuse or pornography – and sexualised photos of kids that will quantity to youngster sexual abuse materials.

The Grok unleashed examine of on-line photos from AI Forensics, based mostly on 50,000 tweets mentioning Grok, revealed between 25 December 2025, and 1 January 2026, discovered that over half (53%) contained people in “minimal apparel”.

The researcher, Paul Bouchaud, reported that 81% of those photos have been people presenting as ladies, and a couple of% depicted individuals showing to be 18 years outdated or youthful, as decided by Google’s Gemini imaginative and prescient mannequin.

The examine additionally discovered that 6% of the generated photos depicted public figures recognized by Gemini. Within the report, Bouchaud mentioned the examine recognized over 350 personalities depicted by the AI picture generator, roughly one-third of whom have been recognized as political figures. Within the report, Bouchaud famous that this implies there are not any obvious guardrails stopping content material that might be used for propaganda materials and disinformation.

Ofcom mentioned it contacted X on 5 January, giving it till 9 January to clarify what steps it has taken to adjust to its duties to guard customers within the UK.

Following the corporate’s response, Ofcom mentioned it has determined to open a proper investigation to ascertain whether or not X has didn’t adjust to its authorized obligations below the On-line Security Act.

Among the many areas of investigation is an evaluation of the danger of individuals within the UK seeing content material that’s unlawful within the UK, and whether or not X is taking acceptable steps to stop folks within the UK from seeing “precedence“ unlawful content material comparable to non-consensual intimate photos.

The investigation can even take a look at how rapidly X takes down unlawful content material when it’s made conscious of it, and the way it’s defending customers from a breach of privateness legal guidelines. Almost about defending youngsters, Ofcom mentioned the investigation can even assess the danger the Grok AI service poses to UK youngsters, and the effectiveness of X to make use of age assurance to guard UK youngsters from seeing pornography.

Commenting on the investigation, Ofcom mentioned: “Reviews of Grok getting used to create and share unlawful non-consensual intimate photos and youngster sexual abuse materials on X have been deeply regarding. Platforms should shield folks within the UK from content material that’s unlawful within the UK, and we received’t hesitate to research the place we suspect firms are failing of their duties, particularly the place there’s a threat of hurt to youngsters.

“We’ll progress this investigation as a matter of the very best precedence, whereas making certain we comply with due course of. Because the UK’s impartial on-line security enforcement company, it’s essential we make certain our investigations are legally strong and pretty determined.”

There was loads of reward for Grok on CEO Elon Musk’s X feed, which seems to have averted discussions on the necessity for guardrails to guard youngsters and folks’s privateness. In a submit on X, Musk mentioned the UK desires “any excuse for censorship”.