The Alan Turing Institute has urged the Home Office to set up an AI Crime Taskforce to coordinate the national response to AI enabled crime.
Its Centre for Emerging Technology and Security (CETaS) has made the recommendation with others in a new report, AI and Serious Online Crime, that says UK law enforcement currently lacks the tools needed to fight an acceleration of AI enabled crime and should adopt a more proactive approach.
This would involve the creation of the taskforce within the National Crime Agency’s (NCA) National Cyber Crime Unit, funded by the Home Office to act as a single point of contact.
It would collate data from law enforcement organisations to identify modes, tools, trends and the geographical diffusion of AI enabled crime. Its work would include mapping bottlenecks in the criminal adoption of AI tools and working with national security and industry partners on strategies to raise barriers to criminal adoption.
The report says that, while the use of AI by criminals remains at an early stage, researchers say there is widespread evidence emerging of a substantial acceleration in AI enabled crime. It is particularly evident in areas like financial crime, child sexual abuse material, phishing and romance scams.
Open-weight exploitation
There are specific concerns regarding the role of Chinese innovation in frontier AI having a significant impact on the threat landscape, with criminals exploiting new open-weight systems – a type of large language model for which the parameters are publicly available – with fewer guardrails to carry out more advanced tasks.
The acceleration in AI enabled crime is being driven by the technology’s ability to automate, augment and rapidly scale the volume of criminal activity, with greater diffusion of AI between the state, private sector and criminal groups leading to more criminal innovation.
Easy exploitation of AI systems and models by criminal groups means these technologies are becoming effective ‘partners’ to criminal groups in achieving their objectives, the report says.
Its other recommendations for ‘proactive disruption’ of AI crime include that law enforcement bodies should rapidly adopt their own AI tools and use technical countermeasures, and that UK organisations should work closely with European and international partners to ensure compatibility in approaches to deterring, disrupting and pursuing criminal groups.
This should come with the setting up of a working group within Europol’s Cybercrime Taskforce focused on AI enabled crime.
Logging and database
Law enforcement should also systematically log the AI tools that are misused for criminal purposes, with the taskforce maintaining a central database.
In regards to testing and guidance, the report says the NCA should produce regular intelligence assessments of the trends in the criminal misuse of AI, and the AI Security Institute should treat fraud domain as an immediate priority for its misuse testing.
Ardi Janjeva, senior research associate at the Alan Turing Institute and an author of the report, said: “As AI tools continue to advance, criminals and fraudsters will exploit them, challenging law enforcement and making it even more difficult for potential victims to distinguish between what’s real and what’s fake. It’s crucial that agencies fighting crime develop effective ways to mitigate this including combatting AI with AI.
Co-author Joe Burton, professor of international security at Lancaster University, said: “AI enabled crime is already causing serious personal and social harm and big financial losses. We need to get serious about our response and give law enforcement the necessary tools to actively disrupt criminal groups. If we don’t, we’re set to see the rapid expansion of criminal use of AI technologies.”
The report is based on research conducted by the project team over a four-month period, involving consultation and interviews with government, industry, and law enforcement in the UK and Europe.
The Alan Turing Institute is the UK’s national institute for data science and AI.
