A Labour MP in the UK has taken Elon Musk’s artificial intelligence company, xAI, to the High Court over claims that its chatbot, Grok, was used to generate fake sexualised images of her.
Jess Asato, MP for Lowestoft, says an AI-generated image made her appear in a bikini without her consent. She has filed the case seeking damages, arguing that the issue goes beyond personal harm and raises a wider question about whether AI companies should be held responsible for how their systems are designed.
The claim, brought under the UK Data Protection Act and for misuse of private information, argues that the harm was made possible by design choices in the AI system, and not merely by the actions of users.
Her legal team says the case is intended to test whether developers of AI tools can be held liable where their systems are capable of producing non-consensual sexualised images of real people.
Following the allegations, xAI reportedly restricted Grok from generating sexualised images of real individuals. However, the MP argues that fixing the system after the fact does not undo the harm already caused.
Strong Public and Political Backing
The case has attracted significant public attention since it was reportedly filed on Wednesday, 3 June. More than 100 campaigners and organisations have issued a joint statement backing the MP’s action, describing it as an “incredibly brave” step towards accountability in AI governance.
Among the supporters are major advocacy organisations including Women’s Aid, Refuge, Rape Crisis England & Wales, the Fawcett Society, the Mental Health Foundation and the Molly Rose Foundation. It is argued that AI systems must be made safe “by design” and that failure to do so has enabled widespread harm, particularly against women and children. The campaigners say the case could open the door to justice for other victims of non-consensual AI-generated imagery.
They also cite research suggesting that between 29 December 2025 and 8 January 2026, Grok generated an estimated three million non-consensual sexualised images of women and children, many of which were later disseminated on X, causing what they describe as “untold harm”.
The case has also attracted political backing. UK Prime Minister Sir Keir Starmer has described the alleged images as “disgusting” and said the MP was right to pursue legal action.
Notably, concerns about Grok had already reached the political sphere before the lawsuit was filed. In January this year, the UK Secretary of State for Science, Innovation and Technology told the House of Commons that AI-generated intimate deepfakes circulating on X were “vile” and, in some instances, amounted to illegal child sexual abuse material. She warned that platforms must act decisively or face regulatory consequences.
The Legal Question: Who Is Responsible for AI Harm?
The legal question emerging from the case is a significant one. Where an AI system generates illegal or harmful sexualised images, should responsibility rest solely with the user, or can it also extend to the company that designed the system?
The claimant argues that AI companies should have a duty to build safer systems from the outset rather than treating safety as a problem to be addressed only after harm has occurred.
If the UK court accepts this argument, the case could mark an important shift in how AI liability is understood, bringing it closer to the product liability model where manufacturers may be held accountable for foreseeable risks arising from the design of their products.
A Ghanaian Lens: What Does the Law Say?
Ghana already has laws that touch on similar harms under the Cybersecurity Act, 2020 (Act 1038). The law criminalises the creation, possession and dissemination of indecent images of children through computer systems and electronic devices, attracting penalties of between five and ten years’ imprisonment under section 62.
Section 67 also criminalises the distribution of intimate images of an identifiable person without consent where there is an expectation of privacy and an intention to cause serious emotional distress.
Taken together, these provisions suggest that Ghanaian law already captures much of the harm described in the UK case, particularly where the victim is identifiable and the images are disseminated without consent.
However, Ghana’s legal framework appears clearer on punishing those who create or distribute such content than on holding AI developers or platform operators responsible for enabling it in the first place. In that respect, the UK litigation presents an opportunity for policymakers to consider whether Ghana’s cyber laws remain adequately equipped for the age of generative artificial intelligence.
Conclusion: Filling the Void in the AI Era
While Ghana’s Cybersecurity Act contains strong protections against intimate image abuse and child sexual exploitation, it does not directly address how liability should be allocated when harmful content is generated by AI systems at scale.
Important questions therefore remain: Should AI developers owe a legal duty to prevent such outputs? Can liability extend beyond individual users to platform operators and system designers? Do existing laws adequately address deepfakes and synthetic sexual imagery that may be generated without any real photograph ever being taken?
As generative AI becomes increasingly sophisticated and accessible, these questions will only become more pressing.
The case against xAI is still in its early stages, but it already has the potential to shape a broader debate about accountability in the AI era.
For Ghana, the legal framework appears to cover many of the harms seen in the case, particularly through strong provisions on intimate image abuse and child protection. However, the question of platform and AI developer responsibility remains less developed.
As AI tools become more powerful and widely used, the challenge for lawmakers goes beyond punishing abuse after it happens. It is about who should be legally responsible for preventing it in the first place.
