Money laundering is a global problem of significant scale. It is believed there is around $4tn of illicit funds circulating at any given time. Regulatory systems have struggled to cope, as the recent Danske scandal has made plain. It is not only the incredible volume but also the speed of financial transactions these days, coupled with a wide web of tax havens spread around the world, centred on the City of London. Despite financial institutions spending around £8bn on compliance in 2017, money laundering is still rife. Some drastic change is needed.
It’s long been good practice for companies to employ hackers to test their systems, or offer rewards for successful breaches. This makes sense, if you want to know how to stop hackers – ask hackers. We need to see an equivalent system or money laundering.
As things stand, anti-money laundering (AML) takes the form of box ticking, of compliance, of following certain routines. It is predictable, static and easy to evade if you know the system. The sort of wealthy criminals who launder money are highly sophisticated and can hire top brains. To avoid being liable banks take a safety first approach, resulting in millions of false positives which all have to be investigated and ticked off. This all strips resources for investigating real incidents.
The launderers know the systems back to front and are continually changing their approach. They are fluid, dynamic and unpredictable. Compliance must be likewise.
Banks need to try a new approach. Instead of targeting transactions, in which the wood can be lost for the trees, regulators must look at the bigger picture. The true target is the money launderers, not the specific transaction. A much more holistic approach is needed.
With recent advances in AI, systems could be made that draw on far more data and context for understanding who the likely launderers are. AI is far superior to humans in finding patterns and leads.
Launderers have clearly beaten humans, no matter how much money and regulation has been thrown at the problem. It is now time to give machines a chance where humans have failed.