The opinions expressed here are those of the authors. They do not necessarily reflect the views or positions of UK Finance or its members.

Anti-financial crime has been at the forefront of these conversations, with AI seen as having the potential to revolutionise how we detect and disrupt the illicit economy. 

The increasing industry interest in AI’s potential to fight financial crime is reflected in academic trends. The scientific research database, Dimensions AI, recorded 5,195 publications related to ‘Anti-Money Laundering’ and ‘Artificial Intelligence’ in 2023, a significant increase from the 2022 figure of 3,964 and the 2021 figure of 2,364. 

Transformation through data

There’s good reason why researchers are more and more focused on AI implementations for anti-financial crime; it’s because AI has proven its ability to do transformative things with large datasets and even larger amounts of unstructured data. 

In a simplified model of the current international financial security system, the bulk of anti-financial crime activity can be boiled down to three main processes. These processes—collection, verification, and interpretation—are performed on ever-increasing amounts of data related to customers and transactions. From the identification and risk assessment of new clients to the detection and reporting of suspicious activity, businesses’ anti-financial crime frameworks depend on the effective performance of these critical data-driven tasks. 

AI’s ability to analyse data and identify statistical relationships—at scales and speeds that far exceed an individual human’s capabilities—means that it has the potential to greatly improve these processes. This is the tangible hope that underpins the hype around new AI financial crime solutions.

A need for change

The need for new solutions might not be apparent to all, but a consideration of the overall effectiveness of the entire anti-financial crime system makes it clear that change is desperately needed. Although more and more resource is invested by businesses to detect and prevent illicit financial activity, a 2023 Europol report estimated that under 2 per cent of organised crime proceeds are recovered every year in the EU. 

Put another way, the estimated $85 billion spent on financial crime compliance in the EMEA region each year yields a return of under $5 billion in recovered criminal assets.

It’s hard to conclude that our current anti-financial crime systems are delivering value for money. While technology alone is never a panacea, AI’s potential to revolutionise the data processes underpinning these systems represents a momentous opportunity for improvement. 

Understanding business needs to shape the future  - The Themis AI in AFC Survey 

The question of how the full potential of AI will be realised in anti-financial crime remains undecided. If we want to see genuine improvements in effectiveness rather than yet more increases in compliance costs, it is critical that the direction of AI developments is shaped by those who best understand what their businesses need from new technology.

This is why Themis has launched its 2025 AI in Anti-Financial Crime Survey, which is gathering the perspectives of senior leaders and decision-makers on how their businesses are approaching the use of AI for their financial crime and compliance functions. The survey will produce a trend analysis report and an accompanying set of recommendations that will be freely available to all respondents (along with access to a free Themis country risk report).

Each response will help shape the future of anti-financial crime; your participation in the survey would be invaluable.