X (formerly Twitter) has announced it will restrict its Grok AI from generating or editing images of real people to show them in revealing or sexualized clothing (such as bikinis or underwear), following intense global backlash over non-consensual AI deepfakes - UKJNews

The change, implemented through technological measures and geo-blocking in jurisdictions where such content is illegal, applies to the Grok account on X and Grok integrated within the platform. It affects all users, including premium subscribers, and comes after weeks of widespread criticism. Users had been prompting Grok (often by tagging it under photos) to "digitally undress" real individuals — including women, celebrities, and in some reported cases, children — leading to thousands of sexualized images flooding the platform.

Elon Musk's xAI and X faced accusations of enabling abuse, with critics highlighting the ease of creating non-consensual intimate imagery. The company has stated it has zero tolerance for child sexual abuse material (CSAM) and illegal content, and users attempting to generate such material face account suspension.

This development follows pressure from governments, advocacy groups, and regulators worldwide, including investigations in California, India, France, Australia, and bans in countries like Indonesia and Malaysia.

What about the UK regulator (Ofcom)?

In the UK, media regulator Ofcom launched a formal investigation into X earlier this month (around January 12, 2026) to determine whether the platform failed to protect users from illegal content under the Online Safety Act. This included assessing risks around non-consensual intimate images, potential CSAM, and whether X adequately prevented or removed such material.

Ofcom has welcomed X's recent changes as a "welcome development," but stressed that its investigation remains ongoing. A UK government spokesperson described the policy update as "vindication" for their pressure on the platform. Prime Minister Sir Keir Starmer had previously condemned the images as "disgusting" and "shameful," warning that X could face consequences — including potential loss of self-regulation rights or stronger legislation — if it failed to act.

A new UK law criminalizing the creation of non-consensual sexual AI deepfakes is also coming into force soon, adding further legal weight.

Here are some key visuals from recent coverage illustrating the controversy and the AI-generated image issue:

Will this be sufficient to prevent action from UK and European regulators?

Not necessarily — at least not immediately. While X's move addresses the core complaint and has been positively noted in some statements, regulators are proceeding cautiously:

• UK (Ofcom): The investigation continues to establish full facts and compliance history. Officials expect robust enforcement, and further measures could still follow if gaps remain (e.g., inconsistencies in safeguards or private Grok usage).
• Europe (European Commission): The Commission has taken note of the restrictions but remains vigilant. It has extended orders for X to preserve Grok-related documents until the end of 2026 amid ongoing scrutiny under the Digital Services Act. Previous condemnations described the content as "unlawful and appalling," and probes in member states (e.g., France) are active.

In short, the policy shift is a significant step that may de-escalate some pressure, but ongoing investigations mean UK and EU regulators are unlikely to fully close the matter without verifying long-term effectiveness and broader compliance. The video you linked (a BBC News report from January 15, 2026) covers this exact announcement and echoes these points on regulatory caution.