Why you should automate your anti-money laundering system

Automated anti-money laundering systems are more secure, but accountants will remain involved in the vetting process.

Why automate AML systems?

Accountants have a duty to ensure their customer due diligence and anti-money laundering (AML) processes are compliant with UK regulations. Vetting all clients’ identities and addresses, and continuing to monitor their behaviour for suspicious activity, is demanding. However, software can do much of the heavy lifting, more robustly and cheaply than via traditional methods.

Indeed, more than that, vetting clients in the traditional way – physically checking passports and proofs of address, for example – is fundamentally less secure than doing so through using software.

According to Ali Jaw FMAAT, co-founder of Severn Accounting, AML systems can monitor in real-time, offering an extra layer of security and assurance. They are more attuned to flag up certain financial transactions or suspicious activities that may be subject to reporting requirements to certain regulatory bodies.

Other key benefits

With more data comes deeper insight. AML software allows accountants to pull together a huge amount of information about their clients. Stuart Hurst, director at Accounts & Legal, says software allows accountants to interpret and be guided by data in decision-making – especially if they’re well acquainted with the client’s trade. Accountants then have two levels of insight: what the data says, and what their expertise tells them.

AML systems can also automatically raise red flags. They surface a great deal of relevant data during the onboarding process and on an ongoing basis. This could include new shareholders, structural changes to the company or a fundamental change in trade. The system will pick it up if a client switches location or points of trade, which should raise professional scepticism.

Gauging risk

Software can tell you if someone is a low, medium or high risk based on the data it finds online. It can be hard to vet international clients, meaning risk can be higher – but software can help with that process.

Apps and insights

New app-based banks, such as Tide, Starling or Monzo also use these checks and balances. To open an account with these banks, prospective customers scan their passports or driving licenses and record a short video in which they say their name. It’s a similar process for large transactions and for forgotten login credentials, with customers recording a new video, which is then compared against the original.

Customers no longer need to go into their bank or accountancy firm with their passport and proof of address. Hurst argues this system is more secure, as users set up their proof of identity directly in the app, and the software cross-references details including addresses and professional history to verify them.

Common issues to avoid

Data gaps can emerge and no matter how advanced the AML software is, it’s only as good as the data available. If some information is out of date or missing that can affect the system’s output, either by throwing up red flags or by clearing the client. It can simply be a case of someone failing to update their address after they’ve moved, but it can still lead to complications where the system’s output is correct based on what is on the public record, and the client’s details have not been updated.

Overreliance on a programme, and insufficient professional scepticism, are also perennial dangers. Automation can lull users into a false sense of security. Interrogating information to avoid false positives is vital.

Jaw had to exercise professional judgement with one client, despite a lack of flags from his AML software. Severn Accounting decided against working with the client because there was the potential for future issues. “They were VAT registered, but they didn’t submit any VAT before they came to us and we were seeing huge transactions with no VAT being accounted for. When we raised it with them, they didn’t engage with us, and we just weren’t comfortable with it.”

Humans have the final say

There are a variety of reasons that a client could be flagged by AML software, though some changes in circumstance may take some time to be picked up by the system. One of those is the introduction of sanctions against a particular country, as with Russia’s invasion of Ukraine.

Accounts & Legal held a review on whether it was practical to carry on with a Russian client. The firm decided not to go ahead even when the system said they’d be cleared, as the timing of a report can affect the system’s output. Hurst explains, “You can run it on the first of the month and it’ll come back approved but if something fundamental changes, it may not be approved when you run it again on fifth of the month.”

Future-proofing with AI and machine learning

As with so many areas of accountancy, the consensus is that artificial intelligence and machine learning will be increasingly adopted in the AML process. The appeal of this stems from vast amounts of data involved in AML compliance, and the increasing complexity of criminal methodologies, which increasingly demand agility in response.

The US-based Financial Action Task Force (FATF) focused on AML AI compliance tools in a 2021 publication on the Opportunities and Challenges of New Technologies for AML. The publication examines AI’s ability to help firms analyse and respond to criminal threats by adding automated speed and accuracy to the compliance process. In particular, it highlighted machine learning as having significant potential due to its ability to learn from data without the need for extensive human intervention.

That human element won’t completely disappear, though, for the reasons above. Hurst is certain that the way the profession is and the way the professional bodies are, they will always want a human being in ultimate control.

AAT Comment offers news and opinion on the world of business and finance from the Association of Accounting Technicians.