The UK ML/TF National Risk Assessment 2025 dropped last week and is a welcome update on those reported in 2015, 2017 and 2020. The NRA runs to 163 pages and provides a summary of the main ML & TF threats and trends since 2020 as well as summarising the most vulnerable sectors and highlights some interesting cross cutting issues.

1. Money Laundering: The main ML related threats are identified as being the “common predicate offences that generate criminal funds” and “harms” since 2020 which come from: “Fraud, Sanctions Evasion, Acquisitive Crime, Drugs, Cybercrime, Organised Immigration Crime (OIC), Tax Evasion, Modern Slavery & Human Trafficking (MS/HT), Online Child Sexual Exploitation & Abuse, Environmental & Bribery & Corruption”. The changing threat landscape reflects the following:

• The UK’s open economy
• The size of the UK’s financial centre and its related professional services
• The use of cash based money laundering remains high despite the decline in the everyday use of cash
• Geopolitical changes including those that have come fro Russias invasion of Ukraine
• The continued vulnerability in some traditional financial and professional services, as well as through UK companies.
• The evolution & adoption of digital finance including non traditional finance including e money and cryptocurrency
• The emergence and availability of AI to both aid both criminal and the response

All regions are vulnerable,

• Urban areas have higher levels of  organised crime and are likely to have elevated levels of money laundering
• Larger cities (London in particular) often prove attractive to criminals seeking cross border and complex ML services due to their concentration of FI/professional services.
• Rural areas can be more susceptible to regional organised crime gangs who predominantly use cash and may exploit locally based professional services to launder criminal funds

The main ML Typologies are Cash, MVTS, Crypto, TBML, Property, Companies & Trusts. Predicate crimes and vulnerable sectors have been mapped to each of these ML typologies, with Drugs involved in all 6, Sanctions, Tax Evasion & OIC to 5, Fraud to 4, Corruption, Waste Management & Acquisitive Crime to 2 & for Sectors: Retail Banking vulnerable in 5, Trust & Company Service Providers & Letting Agents in 4, Wholesale Banking, Wealth Management, MSBs & Casinos in 3, Accounting, Legal & High Value Dealers in 2, E Money and Payment Service Providers and Virtual Asset Services Providers in 1. From an overall sector persecutive (regulated sectors) the following sectors were rated at high ML risk (residual risk = inherent risk minus mitigation):

• Banking (Retail, Wholesale (including Markets) and Wealth Management), MSBs, Legal, Accountants, TCSP and VASPs. This represents no change from 2020 except for VASPs where the rating has increased to High from Medium.

2. Terrorism/Terror Finance: The UK terror threat level remains at “Substantial” in the middle of its 5 point scale. Terror ideology affecting the U.K. is made up of 3 main types, Islamic related, far right related and terrorism from Northern Ireland. 5 types of terror groups are assessed, namely lone actors, small groups. Overseas groups, decentralised groups and groups effectively holding territory. Their financing needs and modus operandi differ. From an overall sector persecutive (regulated sectors) the following sectors were rated at high TF risk (residual risk = inherent risk minus mitigation):

• Retail Banking, MSB’s & EMI/PSP. This represents no change from 2020 except for EMI/PSP where the rating has increased to High from Medium.

3. Cross Cutting Issues – A number of cross cutting issues have been identified namely: 

• Artificial Intelligence
• Schools & Universities
• Football clubs and Football Agents

3.1 Artificial Intelligence

Can be used for Ill as follows:

• Use of AI for synthetic bank account creation, fraud and impersonation, phishing and onboarding of money mules,
• Use of AI in money mulling, by automating mule herding activities,
• Use of AI to lower the bar for novice criminals to create bank accounts using synthetic data
• Evading and defeating AML detection systems 

For Good as follows:

• To reduce admin tasks to free up time for law enforcement
• To reduce false positives in the private sector
• AI enhanced behavioural analytics may improve detection including increased data sets, and connecting customer transaction, social media and other open source data
• To counteract deep fakes and fraudulent documents

3.2 Schools & Universities

Many UK schools & universities have international reputations that attract not a only academic scholars but also criminal funds designed to support the education of criminal family members including foreign PEPs. Risks if laundering corrupt funds, or funds subject to sanctions & or connected with other forms of criminality is a concern. Students can also become attractive to criminals especially those that want to recruits money mules for payments.

3.3 Football Clubs & Agents

Football Clubs and Agents may be targets for criminality especially clubs that are financially distresses most likely lower down the UK pyramid. Alongside the potential for money laundering, illegal betting, match fixing and other fraudulent activities, including bribery may be crimes that could be visited on vulnerable football clubs including by the aid of unprofessional agents.

4. Comments and Final Remarks

The NRA is an important document that forms the basis of a countries collective understanding of it’s ML & TF risks, from which much flows. Despite 5 years passing since the last UK ML/TF NRA much has changed since then, but the threats & vulnerabilities are largely unchanged, with a few exceptions. These changes are nevertheless uncontroversial & are already well understood & likely catered for, at least by proactive risk managers. The main changes relate to:

• Electronic Money and PSPs and VASPs which have had their ML risk ratings increased & EMI/PSPs their TF risk ratings increased to High.
• Football clubs and Agents, Schools and Universities and Artificial Intelligence called out as “cross cutting issues” with heightened awareness needed. AI is highlighted as it can be used for good and ill and both are explored based on a current understanding and future potential impacts. One area that appears to be open for further data collection  and analysis using AI is social media monitoring, but before any such approach is taken, this should help considered carefully as the regulated sector including Banks should not ordinarily be monitoring customers and prospects private and semi private or even senior public lives.

Whilst the main threats are generated based on likely proceeds of crime generated, estimates for each & in aggregate are largely absent, which suggests intelligence & confidence levels on prevalence and proceeds is low, as it has been for some time. A better model than focussing on proceeds, is a combined assessment estimating harms, prevalence, proceeds and costs (with weighting towards proceeds for a ML assessment & towards harms for a TF assessment).

The assessment focusses on the threats and vulnerabilities, which makes sense. The sectors included though are only those that are within the system and regulated. The addition of cross cutting issues is helpful, which broadens the focus. Nevertheless, with fraud at 43% of national crime figures, where are the fraud facilitators, the Tech, the Digital, Social Media, Online Platform Enablers that facilitate fraud. Whilst they don’t facilitate ML, they do facilitate (mostly unwittingly) the crime, so they are not included. That is a mistake as ML is a tool to tackle the crimes and they could have been added as cross cutting issues, at least. The same can be said of MS/HT sectors, such as Massage Parlours, Nail Bars and Hand Car Washes, Fruit Picking & Deep Sea Fishing, IWT sectors, such as the international pet trade. Another important and missed cross cutting issue is Brexit (for more see below).

Whilst the UK NRA is a very useful and a timely publication, the overall headline assessment is that ML risk is rated at “High” in 2025. It nevertheless doesn’t answer a very important question, which is whether the UK’s inherent ML risk and or the UK’s actions to mitigate the risk has gone up or down?

On the face of it, the inherent ML risk should have been positively impacted by the UK’s reaction to the Russian invasion of Ukraine, though e.g, due to Brexit, UK companies & businesses have had to look for new markets beyond the EU, where bribery & corruption & criminality are more a problem. The inherent risk has probably been more adversely impacted by the risk in fraud & scams. Still mitigation should be improved due to the many efforts & initiatives of the public & private sectors since 2020 against fraud & scams but also more broadly in the PPP, and benefited from the economic crime levy. An own goal again comes from Brexit which will have impacted cooperation between the UK & EU. These mitigation actions, whilst significant have yet to translate into genuine measurable material effectiveness gains such as money laundering prosecutions & convictions & asset recoveries, despite a number of important victories. Nevertheless, despite significant investments in asset recoveries, the UK has managed to recover just £358 million in 2021/22, £241.5 million in 2022/23 & 243.3 million in 2023/24 (source NRA 2025). Whatever the estimate of proceeds of crime, this is likely to remain well below 1% of likely proceeds of crime. Other countries have increased their asset recovery levels to at or around 5% during the same period, including Italy, Singapore, Latvia & the UAE, which shows it can be done, and more is needed in the UK to convert action into results.

Financial Crime News – 21 July 2020

For this Article and the 2 Page Summary Dashboard: See: HERE UK NRA 2025