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

Read my previous blog here.

The question of how organisations are actually adopting AI to protect themselves and society from illicit finance threats is an important and timely one. To answer it, global financial crime specialist Themis conducted an in-depth survey of 74 senior decision-makers across industries and jurisdictions.

The findings, published in our report AI in Anti-Financial Crime: The State of Adoption in 2025, reveal three key insights:

1. A wave of AI adoption Is coming

Currently, 69 per cent of respondents are not using AI in their Anti-Financial Crime (AFC) or compliance systems. However, this is set to change rapidly – 51 per cent of non-users plan to adopt AI within three years and over 80 per cent expect to have AI-driven AFC systems by 2030. 

Legacy technology remains a major hurdle – 39 per cent cited it as a significant obstacle. Yet, among those already using AI, tools are applied across eight distinct AFC functions. These range from KYC and due diligence to customer risk scoring and investigations. This suggests that AI adoption has the potential to be broad and cross-functional across AFC practices – an approach to organisational AI uptake seen as best practice by the Harvard Business Review. 

2. Senior leaders are ready 

Encouragingly, 85 per cent of respondents believe their leadership has at least an intermediate understanding of AI’s risks and opportunities, with 50 per cent rating this understanding as high. 

The top motivation for adopting AI? Efficiency – leaders see AI as a way to save time and cut costs, especially in overcoming outdated systems. 

But it’s not just about improvement. Leaders also recognise AI’s potential to address emerging threats. The second most-cited reason for adoption was to “deal with new types of risk” – many of which are, themselves, AI-powered. 

3. Barriers to adoption exist - but can be overcome 

Despite optimism, organisations still face hurdles. These fall into three categories: 

• Technology concerns: Reliability and accuracy were the top worries, especially given the sensitive nature of AFC processes. These concerns are, of course, valid, and should remain present in any organisation which sees AFC as more than a box-ticking exercise, but they are being allayed by global efforts to understand and enhance the trustworthiness of new AI systems.  

• Internal challenges: 45 per cent cited knowledge gaps among staff as a major obstacle to adoption– far more than the 16 per cent who expressed general scepticism about AI. We see this as a positive finding for organisations keen to adopt AI – there is no shortage of training and capacity building available, but cultural change can be far more challenging to embed  

• Uncertainty about regulation: Over a third of respondents flagged this as a concern. While AI regulation is still evolving, the UK regulatory principles clearly outline the areas to be considered, while regulators like the FCA have driven forward initiatives to propel the development and adoption of new AI tools. 

As the technology and regulatory landscapes continue to mature, many of the barriers to AI adoption will diminish. Familiarity with AI tools and a clear understanding of their applications will be key to successful adoption. 

What’s next? 

Our full report offers six actionable recommendations for organisations looking to integrate AI into their AFC strategies.  

Now is the time to move from awareness to action. The tools are here. The leadership is ready. The risks are evolving. The question is no longer “if” AI will transform anti-financial crime efforts – but “how soon?.”

This research was made possible with the support of UK Finance. We thank them for supporting the dissemination of the survey.