Has AI gone too far? UK policy amendments to be made to combat deepfakes made by X’s AI Grok

Grok AI is a chatbot developed by xAi, on ‘X,’ created by Elon Musk. It differentiates itself by giving funny, witty responses, making it different from other LLMs (large language models) like ChatGPT as it displays a more humanistic feel. The word ‘Grok’ itself is inspired by an element of Robert Heinlein’s 1961 sci-fi novel Stranger in a Strange Land; to encompass concepts like love, hate, fear, and shared reality— “to understand so thoroughly that the observer becomes a part of the observed.”

However, following recent events, the EU Commission announced that it would examine whether the tool used on X met its legal obligations under the EU digital services act after thousands of disturbing searches involving children, amongst other sexually exploitative images of women, came to light. Such images are generated by using ‘prompts’ to instruct Grok to produce images and videos of women and children in bikinis, underwear, or fully nude. 

A researcher at Bloomberg found that users were generating over 6,700 of photos of this nature every hour. Ashley Sinclair – mother of one of Elon Musk’s children – has recently spoken out, filing a lawsuit alleging that Musks’ Grok has been used to generate dozens of non-consensual, sexually explicit photos of her, some from when she was just 14 years old.

Countries such as Indonesia and Malaysia have prohibited use of the chatbot, stating that the chatbot lacked adequate safeguards to prevent the generation of “obscene, indecent, and grossly offensive” content, and other European countries, including the UK, are contemplating doing the same. Elon Musk has spoken out on the matter, stating that the UK will look for any excuse for censorship.

The UK has launched an Ofcom investigation, which recalls that UK legislation passed in 2025 criminalised the creation or request of intimate deepfakes without consent, but has yet to be fully implemented. They call on the government to set out what enforcement action will be taken, when the law will be brought fully into force, and what immediate steps will be taken to protect users, particularly women and children, from further harm caused by AI-generated sexual abuse. However, this is not enough, and the response and lack of action from the government is ultimately humiliating to the victims who have been sexually exploited in such a derogatory way. 

Under the UK online safety act, this is a matter of criminality, where generating child abuse is a crime. Nevertheless, the nature of such violations makes it extremely difficult to convict the sex offenders. Unfortunately, this type of offence is not the first of its kind. Such horrific crimes involving AI are only getting more and more prevalent, with countless other ongoing investigations such as Google’s AI tools being used to subject women in public life to online violence and even murder. 

Such harrowing imperilments demonstrate that activities once confined to the dark corners of the internet have entered the mainstream at a concerning scale, endangering women and children in a way never seen before. The era of AI-assisted online violence – particularly targeting women and children – is no longer looming; it has arrived. And it is worsening the threat landscape for women in both public and private settings.

Words by Lily Rodney