As losses from financial crime continue to rise, AI is often portrayed as a way to make compliance teams more efficient and effective — but how well is it really being adopted?
Napier AI has released the second annual iteration of the Napier AI / AML Index, which can be downloaded for free. This offers an in-depth global ranking of AI’s impact on anti-money laundering (AML). A total of 40 countries, spanning Europe, North and Latin America, Middle East, Africa and Asia Pacific, have been ranked based on their effectiveness in financial crime compliance.
The report was compiled by data scientists who conducted extensive research, as well as interviewed senior compliance executives at 25 financial institutions around the world.
Speaking to FinTech Global about the release of the report, Janet Bastiman, Chief Data Scientist at Napier AI, stated, “The Index is designed to pinpoint what good looks like in the careful balance between efficiency and effectiveness in anti-money laundering. The Index, like Napier AI itself, is designed by data scientists (including me) and built on data science. It is designed to illustrate the art of the possible for AML and AI, by highlighting national regimes and financial services markets that are achieving great AML outcomes through innovations in AI.”
How countries were ranked
Countries are given an overall index score, and a lower score indicates better performance. However, the report also gives each country four additional individual category scores.
- The first of these is the AML attitude score. This is based on the sentiment analysis and scoring of expert interviews. This covers responses on attitude towards AI, money laundering growth, alerts, identification and prevention compared to the previous year.
- Secondly, there is an AI/AML regulation score, which is based on responses around AI usage and the role of regulation in the region. This score is influenced by whether respondents believed regulation was a help or hindrance to AML outcomes and the impact on compliance costs.
- Score three is the total cost of compliance (TCO). This is the ratio between total compliance costs in each region and the region’s money-laundering losses. The cost is based on interviews with compliance leaders. Each country’s score reflects how closely its ratio of compliance spending to money-laundering losses aligns with the ideal range identified by the study.
- Score four is AML effectiveness. This is based on the estimated money laundering losses as a ratio of GDP. It puts the total value of estimated losses to money laundering into context and provides a fair comparison between national economies.
Bastiman added, “This range is lower in 2025’s study than 2024’s, owing to stronger money laundering prevention outcomes among the top-ranked countries. As a result, many cost of compliance scores have changed significantly from last year.”
The final score is the AML effectiveness score. This is based on the estimated money laundering losses as a ratio of the GDP of the country. This score puts the total estimated losses from money laundering into context and allows fairer comparison between national economies. Total losses were modelled with inputs from the Organisation for Economic Co-operation and Development (OECD), the United Nations Office on Drugs and Crime (UNODC), The Basel Index and the Financial Action Task Force (FATF) Technical Compliance and Effectiveness assessments, and more.
“It is worth noting that we modelled based on 80% of total money-laundering losses, so these are conservative estimates with actuals potentially much higher,” Bastiman added.
Biggest trends from the report
The United Nations Office on Drugs and Crime (UNODC) estimates that between $800bn and $2trn is laundered annually, which makes up around 2% – 5% of the global GDP. However, it is not just the losses that are rising, so is the cost of compliance.
AI will not only reduce manual workloads but could also help firms save billions by improving AML workloads. Bastiman noted, “The Index found that regulated firms like banks, payments firms, wealth and asset managers, telcos and insurance companies can save $183bn (up from $138bn last year) on compliance costs by implementing AI into their AML strategies. And $3.3trn could be returned to global economies with AI-powered AML strategies (up from $3.13trn last year).”
With such a huge opportunity available, Bastiman emphasised that using AI for AML is no longer a niche area. Instead, it is being eagerly embraced, with numerous national initiatives underway to support AI use cases.
While all countries can embrace AI to improve AML, the Napier AI / AML Index spotlights the countries that have the greatest opportunities for AI-driven cost savings. At the top spot is the United States, with savings of $26.1bn. It is followed by Germany with $14.3bn and France with $11.1bn in savings.
An interesting case rests with Australia, which has made progress in AML efficiency and effectiveness. As such, Napier AI highlights that it doesn’t have huge inefficiencies that can be addressed with AI.
While the country has been working hard to improve the effectiveness of AML, the country ranked the lowest in the AML attitude category.
Bastiman explained, “Australia remains near the bottom of the AML attitude rankings for the second year in a row and also continues to score poorly for AI/AML Regulation, perhaps indicating the negative sentiment of the market regarding the ongoing AML reforms creating significant challenges for FIs, but ultimately demanding excellence in AML from its market participants. The real-terms improvement in its AI/AML Regulation score could also indicate that while AML reforms have caused some pain, ultimately the market believes that they are directionally correct, and see the potential for AI to remove some of the operational burden.”
On the other hand, the UK and US have ranked much better in their AML attitude scores this year, which Bastiman suggested could be a result of fatigue from rising financial crime, alert volumes and sanctions inflation. The most impressive change comes from New Zealand. The country was ranked in last place last year and has risen to first place in the new report. “Interestingly New Zealand ranks amongst the worst markets when it comes to AI/AML regulation score, indicating that regulatory enablers aren’t keeping pace with FI appetite.”
To highlight some of the other findings, the countries with the highest TCO were Poland, France and Germany, which have maintained similar positions on last year. Singapore, the UK and Italy continue to be top performers in AI/AML regulation scores, while France, Germany and Poland showed significant YoY improvements.
Bastiman added, “For every market evaluated within the Index, there’s an interesting story to uncover amongst the four category scores, even for those that rank lowest in certain categories (such as AML Attitude). These indicate the reasons for challenges in AML in the region, as well as providing some forecast for when the tide might turn in favour of the financial institutions fighting the good fight.”
Four key themes impacting financial crime professionals
Bastiman noted the report uncovered four main themes that resonated across all financial crime professionals. These were transforming compliance from a cost centre into a competitive advantage, building trust into AI, the challenges of divergent regulations for AI and ballooning costs of compliance.
When it comes to compliance costs, the report offers an interesting perspective into the problem for countries. For instance, it provides a ratio for the spend on AML versus the amount of money laundered in the context of GDP.
The ‘effective leaders’ in this category are the ones that are losing less than the global average of 5% GDP to money laundering and are not overspending on AML. These include the Nordics, Canada and Spain. At the other end of the spectrum are countries like France and Poland that spend heavily on AML compared to the size of the problem. Bastiman noted, “These countries have a lot of space to benefit from AI for increasing effectiveness and driving down the total cost of compliance.”
Another theme highlighted was changing the perspective of compliance. It is common for compliance to just be viewed as a cost centre, a tick-box exercise that needs to be done. However, this is an outdated viewpoint and embracing compliance through technology can uncover a variety of competitive advantages.
That said, Bastiman emphasised that achieving this is not a case of simply investing in AI. “To turn compliance from a cost centre into a key capability, it requires balancing innovation with the right governance and oversight.” The index estimates the ‘ideal’ ratio of compliance spend to money laundered is between 1.36% and 3.36%.
“We believe that spending either within – or close to – this range will provide good money laundering mitigation outcomes, while also maintaining operational efficiency of compliance departments. Those that sit within this ideal range, exhibit a good balance of strong understanding of underlying risk, robust regulation and enforcement, and effective adoption of AI.”
For more insights into the current state of AI in AML, download the free report here.
Read the daily FinTech news
Copyright © 2025 FinTech Global
Copyright © 2018 RegTech Analyst
