Michael Howard 23 Dec 19 7 min read

Fighting financial crime: AI versus money launderers & Where's Wally?

Where’s Wally? A great question asked of children around the world by the British illustrator Martin Handford. Spending minutes or hours engrossed in these books, children would cast their eye over each page, with varying levels of scrutiny, to find a man and his dog who were highly skilled at blending in amongst a sea of red and white.

While we have no significant back story to work with, Wally (or Waldo, Walter, Willy, Charlie, or Efi as he was referred to in other parts of the world) was a man of mystery, who turned up in random scenarios and places around the globe.

There’s a surprising similarity between Where’s Wally and the task faced by anti-money laundering compliance officers, who monitor dealings and transactions within their business for suspicious people who, paradoxically, often fit in yet markedly stand out. While Wally is confined to the pages of a book, money launderers and other financial criminals are unrestrained with their whereabouts. These money launderers are also getting smarter at using digital mediums to move money around. And this is where AI enters the fray and becomes the game-changer in the fight against financial crime.


What is AI?

Artificial intelligence isn’t just a piece of software or technology – but an ecosystem’s ability to make decisions and carry out work for, but independently of, a human, and with an increasing success rate. The three important aspects here are ecosystems, independence, and increasing success. As far as the ecosystem is concerned, it’s the bringing together of various technologies to deliver intelligent outcomes for people, not just streamlining processes. Regarding independence, AI can do the bulk of the heavy lifting or processing of information to make people’s lives easier. Finally, on increasing success, AI needs to have a framework to work and learn from, so it knows how to make correct decisions with increasing accuracy.

For example, if you were to use AI with the Where’s Wally books, it could be trained to use computer vision to spot Wally in a fraction of a second. And not just on each page but on every page of every book that has ever been produced. Taking approximately eight weeks to illustrate each page, who knows what the British author would have thought about the kind of scale and speed that AI can solve tasks like this.

So, AI can be used to spot characters in a book, but what about in the fight against financial crime?


How does AI work with AML?

One of the hardest things for criminals to do is get dirty money into the system. Once they do, that money can be sent around the world, in and out of safe havens and cryptocurrencies, obscuring its sordid past. Or so one thought. With AI, it’s highly possible to spot suspicious activity, and it can do this in a few ways.

Transaction monitoring.
You’d be surprised at the lengths people do (or don’t) go to to try and make themselves look inconspicuous when moving money around a financial system. The way AI combats this is to process large volumes of transactions, which builds up profiles based on the patterns it sees for each customer. The AI is trained to spot transactions that fall outside of these patterns, and can automatically flag or produce a suspicious activity report for each occasion.

Machine learning.
Using machine learning, the AI is trained to reduce false positives, which minimises the workload of an AML compliance officer.

Shadow profiles.
Smart anti-money laundering software doesn’t necessarily need to be account-based either, which is beneficial for businesses such as foreign exchange shops. This is because when transactions are below a certain figure, their customers can complete the currency swap anonymously. The AI can build patterns of such people’s behaviour and flag particular transactions, so when a similar one presents itself in the future, businesses can be better placed to respond accordingly.

Watchlists and KYC.
Wally’s multiple identities across different jurisdictions would make the AML compliance officer’s task even harder. This is where another aspect of the ‘ecosystem’ comes to the fore. Watchlist screening and knowing your customer is a huge aspect of AML compliance as well, with businesses servicing on those tasks alone. What makes this even easier is when supervisors from other jurisdictions work together, sharing and connecting data sets from their part of the world. The FBI, for instance, acknowledge that criminals have little concern for geopolitical boundaries, and are testing an AML partnership with the Australian government in the fight against financial crime.

With criminals getting more and more sophisticated, businesses need to keep investing in this space as well or they will put their customers at risk and be at risk of brunting the might of the financial supervisors.


Where to with Wally?

As we’ve just read, criminals who thought they were dealing in the shadows have practically no chance of avoiding the watchful eye of an AI-powered AML system. Like the protagonist in Where’s Wally, it’s only a matter of time before criminals are caught and turned over to the supervisors for further investigation. And that time is getting shorter and shorter.

So the question for you now stands, are you finally ready to put aside your manual AML practices and let AI make meeting your compliance obligations exponentially easier? If you answered yes, then you might want to consider reading this ebook on automating your AML compliance.



Ready for a demo?