Combating the Global Economic Impact of Money Laundering Using Graph Technology
- Blog >
- Combating the Global Economic Impact of Money Laundering Using Graph Technology
Money laundering is an important concern for many companies, especially in financial services companies. Employees at these companies are required to undergo fundamental training on identifying signs of money laundering, regardless of their position or seniority within the company.
Money laundering is simply just that prevalent and costly to these companies; in many cases, an anti-money laundering compliance program is compulsory.
During their presentation, Making Anti-money Laundering Effective, at Graph + AI Summit last week, Sopra Steria’s Andreas Vermeulen and HSBC’s Bart Visscher shared some eye-opening statistics that emphasized just how costly money laundering can be.
Money Laundering’s Noticeable Global Financial Impact
The estimated amount of money laundered in a given year is staggering. According to the UN Office on Drugs and Crime, between 2-5% of global GDP is laundered each year. That number equates to somewhere between $800 billion and $2 trillion annually.
Putting that into perspective, according to the presentation, that number equals the total global agricultural output for the year. That’s an incredible amount of money in the hands of criminals, but it’s not the only impact money laundering has on affected organizations.
In 2020, the UK Financial Intelligence Unit (UKFIU) received and processed almost 600,000 suspicious activity reports (SARs). These reports are generated when, as the name suggests, suspicious activity is detected and requires further investigation. Over 95% of these reports came from financial institutions, costing them a staggering total of £28.7 billion.
If you break it down, each SAR costs the financial institution reporting it about £56,000. With that kind of investment, you’d imagine these institutions would have some success preventing money laundering.
Unfortunately, that’s not the case – these anti-money laundering measures prevented a mere £171 million; for every £1 invested, only 0.6 pence of money laundering was prevented. It’s a huge, complicated problem to overcome.
“Out of the tens of millions (of alerts) we end up in the hundreds of thousands of SAR reports,” said Bart Visscher at HSBC. “Out of these hundreds of thousands of reports we send to the financial crime division, we may get a few thousand prosecutions. So we’re going from billions to thousands. That’s really looking for a needle in a haystack in a field of haystacks.”
Building a Better Anti-money Laundering Mousetrap
The vast amount of money laundering taking place in the world is just one of the roadblocks. Because money laundering is so profitable, those committing these acts are incentivized to be innovative.
“There’s a ‘cold war’ of ideas happening,” according to Bart. “As we come up with new ways of detecting money laundering, money launderers come up with new and better ways to disguise it, and it takes sometimes years for us to find these new ways.”
Complicating matters further, privacy regulations limit the amount of information sharing between organizations, which also limits the ability of organizations to train their machine learning solutions. That’s not to say that data privacy isn’t important; it’s merely pointing out that these measures do make anti-money laundering efforts more difficult.
Graph technology can be part of the solution, something the presenters illustrated during the second half of their presentation. Beginning with federated learning to address data privacy concerns and a distributed ledger to share machine learning capabilities, the presenters showed that improvements to anti-money laundering capabilities are not impossible.
“For the distributed ledger to properly work, this is where we really get some advantages out of using extensive graph theory and graph networks,” according to Andreas Vermeulen at Sopria Steria, who later added, “We need to create a hyper-scaled, federated learning network that is sharing information between graph networks and actually enabling us to create a virtual graph network.”
Money laundering is of vital concern, not just to affected companies, but as an impactful criminal activity. “It is not so much a white-collar crime,” concluded Andreas. “This is horrible stuff that happens, and we can use technology to actually fix it.”
You can learn more about Bart and Andreas’ recommendations to make anti-money laundering more effective by watching the full Graph + AI Summit session.