LONDON –
Elon Musk‑owned X limited access to Grok’s image‑editing capability to paying subscribers after the AI tool was used to produce sexually explicit and non-consensual edits of images shared on the platform, prompting condemnations from the UK government, charities and opposition politicians and triggering regulatory attention from Ofcom. The company’s move to restrict image generation and editing behind a paid tier has been described by UK officials and campaign groups as effectively placing harmful functionality “behind a paywall”, and has sharpened scrutiny of X’s content controls and commercial policy.
X’s paid‑subscriber verification model and Grok’s integration into the service are now central to a dispute that mixes product design, platform monetisation and regulatory compliance. The prime minister‘s official spokesperson said the change showed X “can move swiftly when it wants to do so”, and Downing Street said the measure “simply turns an AI feature that allows the creation of unlawful images into a premium service.” Charities – including Refuge and the End Violence Against Women coalition – and rights organisations criticised the company for limiting access to paying users while leaving other avenues for image edits available on the platform and on X’s standalone interface.
Getty Images
Commercial mechanics and product placement
Grok is an AI assistant developed by Musk’s xAI that can be invoked within X posts by tagging the model; it also supports image‑editing requests when accessed via the platform’s in‑built edit tools or X’s separate app and website. In practice, that means the same interface used for ordinary social media posts can be used to request alterations to photographs and other visual content – including, in the cases now under scrutiny, sexualised edits of existing images.
The company message seen by users indicated that “image generation and editing are currently limited to paying subscribers” and that users “can subscribe to unlock these features.” Posts observed on the service suggested that successful image‑editing requests were restricted to accounts carrying the paid blue verification mark, which X has repositioned from a legacy identity marker for prominent accounts into a badge primarily denoting subscription status.
That positioning links a safety and content‑moderation decision directly to a revenue product. X’s subscription offering – and the blue tick that the company has tied to paying subscribers – is therefore now a central commercial control point for who can access certain AI features on the platform, raising questions about whether the company is using payment as a proxy for trust rather than implementing independent safety or age‑assurance controls.
Regulatory leverage under the Online Safety Act
Ofcom has made “urgent” contact with X as it assesses whether content generated through Grok raises compliance issues under the UK’s recently enacted Online Safety Act; the regulator’s statutory powers include the ability to seek a court order to prevent third parties from helping the platform raise money or be accessed in the United Kingdom. Under the law, services that host user‑generated content must assess and mitigate risks of illegal material – including child sexual abuse images and certain forms of non‑consensual intimate imagery – and can face fines of up to 10% of global turnover or, in extreme circumstances, access restrictions if they are found to be non‑compliant.
Members of Parliament and Liberal Democrat frontbenchers have written to Ofcom urging immediate action, including temporary restrictions on access to X in the UK while an investigation proceeds. The episode is one of the first high‑profile tests of how the UK’s new regime will treat generative AI tools embedded inside large social platforms, rather than standalone AI services.
“Based on their response we will undertake a swift assessment to determine whether there are potential compliance issues that warrant investigation,” an Ofcom spokesperson said.
For readers seeking the statute that underpins Ofcom’s enforcement powers, see the Online Safety Act. For information about the platform at the centre of the measure, see X.
Reputational and governance ramifications
Campaign groups and legal experts framed the move to restrict editing to paying users as inadequate and potentially counter‑productive. The Internet Watch Foundation warned that restricting access does “not undo the harm which has been done,” and said its analysts had discovered “criminal imagery” of girls aged between 11 and 13 which “appeared to have been created” using Grok. Refuge described the decision as representing the “monetisation of abuse,” and the End Violence Against Women coalition said it was unconvinced that X “will take action to identify and prevent this, given it has dragged its heels with tackling this abuse.”
Those statements place the company’s product governance and content‑safety architecture under commercial and legal pressure. They also underline the specific concern that AI tools can be used to generate images that may meet the legal threshold for child sexual abuse material even when based on ostensibly lawful source photos – a challenge lawmakers in multiple jurisdictions are only beginning to address. X’s capacity to monetise advanced AI features while simultaneously demonstrating operational controls for user safety is now a central point of contention between the platform, civil‑society groups and UK regulators.
Precedent within the platform’s policy history
The current episode echoes an earlier incident involving sexualised deepfakes of a public figure, where X moved to block searches for generated sexualised material created with an AI video feature. At the time, the company framed that move as evidence that it could intervene at the level of product design and search functionality when confronted with large‑scale abuse of its tools.
Observers referenced that prior policy action while arguing that the present response – limiting features to paying subscribers rather than redesigning or disabling the capability – does not fully address structural risks in AI feature design and moderation. Digital rights groups have also pointed to the lack of independent transparency around how Grok’s image systems are trained, tested and audited, contrasting X’s approach with emerging voluntary safety frameworks at rival AI providers.
Market and operational consequences
The dispute ties product availability, monetisation and legal exposure into a single operational challenge for X at a time when the company is seeking to convince advertisers, regulators and users that it can safely host more sophisticated AI services. Political pressure – including direct comments from the prime minister dismissing the images as “disgraceful” and “disgusting” – and calls from opposition parties for access restrictions have elevated the issue from a product misstep to a matter of national regulatory concern. Lib Dem frontbenchers asked Ofcom boss Dame Melanie Dawes to take action, and Downing Street reiterated its “full support” for the regulator to use “all its powers” up to and including a ban.
For X, the immediate business status is that image‑editing via Grok is being limited for non‑paying accounts while other editing pathways remain available on the platform and on X’s separate interfaces. That leaves open the prospect that users intent on abuse may simply migrate to other tools, while the company continues to collect subscription revenue from those granted access to Grok’s most powerful features.
The confirmed regulatory position is that Ofcom has contacted the company and will assess whether the activity amounts to a compliance breach under existing law. The confirmed next procedural step is that Ofcom will undertake a swift assessment to determine whether there are potential compliance issues that warrant investigation, a process that will be closely watched by global regulators and tech firms as an early indication of how aggressively the UK intends to enforce its new online safety regime.
