Sean Gould
Researchers are raising concerns over the rapid rise in AI chatbot gender-based violence. A recent report by Durham University, University of Lucerne and Swansea University warns that chatbots are now actively reshaping violence against women. Creating “dangerous blind spots” in the UK’s proposed policing legislation that urgently need addressing.
The report, Invisible No More: How AI Chatbots are Reshaping Violence Against Women and Girls, identifies three distinct ways in which chatbots are contributing to violence against women and girls (VAWG). It argues that advances in AI technology have not only undermined female safety but also created entirely new forms of abuse that existing UK legislation was never designed to address.
Types of chatbot gender-based violence
The report highlights a new and deeply concerning form of abuse known as chatbot-driven VAWG. Researchers examined cases where AI chatbots autonomously perpetuated gendered violence without direct prompting from users.
According to the study, AI chatbots are now capable of generating image-based abuse and replicating behaviours inspired by harmful online subcultures.
The researchers stated:
“In naming the chatbot as the driver and sometimes perpetrator, we are not absolving developers or users of responsibility. We are highlighting that this is a new way that abuse is being committed, and the chatbot is playing the central role in the act, sometimes in ways not wholly envisaged by design choices.”
Chatbot-enabled violence against women and girls
The report also identifies chatbot-enabled VAWG, where chatbots provide detailed advice and guidance to users seeking assistance with stalking, grooming and child sexual abuse-related behaviour.
Researchers warned that although chatbots may not directly commit the abuse, they can equip users with the information required to carry out gendered harm. This creates significant legal and ethical difficulties around attributing responsibility between the developer, platform and end user.
Chatbot-simulated violence against women and girls
The third category identified is chatbot-simulated VAWG, described as a new and distinct abusive practice in which chatbots generate violent gendered scripts designed to target women.
Although these simulations are initiated by humans, the report warns that such roleplay scenarios risk desensitising users and normalising abusive attitudes towards women and girls.
What can lawyers do when AI chatbots produce harmful content?
The current UK legal framework was not designed with autonomous or semi-autonomous systems in mind. As a result, there is currently no single, coherent cause of action for “AI harm”. Lawyers are therefore required to fit these emerging forms of misconduct into existing legal doctrines.
This creates several challenges. Questions often arise around attribution of responsibility, foreseeability and evidential thresholds. Much of the harm caused by AI systems, particularly where psychological manipulation, synthetic media or automated harassment are involved, does not fit neatly within traditional categories of liability.
Potential legal remedies
There are, however, legal avenues available.
The UK GDPR and the Data Protection Act 2018 may provide a route where personal data has been misused. This may include cases involving deepfake imagery or the scraping and processing of identifiable information.
Negligence claims may also arise where a developer or platform has failed to implement reasonable safeguards against foreseeable harm. However, these claims are often complex and highly fact-sensitive. They are also not always well-suited to dealing with urgent or ongoing harm.
In practice, the most effective remedies are likely to arise through existing harassment, abuse and image-based sexual harm legislation. These provisions provide both criminal and civil protections.
Statutes such as the Protection from Harassment Act 1997, alongside newer offences criminalising the creation and distribution of non-consensual intimate images, including AI-generated content, offer more direct protection. These frameworks allow courts to grant injunctions, prosecutions and ancillary orders capable of quickly disrupting harmful conduct.
If you are involved in a case involving Ai chatbot gender‑based violence or any other form of gender-based or domestic abuse and would like to discuss possible courses of action with one of our experienced criminal barristers in Brighton, please contact our clerks today.
For specialist support with website design, development and digital presence, please visit the Simdure website team