Despite widespread advances in anti-money laundering technology, experts doubt whether it will ever be able to match the capacity of criminals.
There is even a belief that technology may only ever be as good as five to ten years behind what criminals can achieve at any point, a conference was told.
Matthew Leaney, CRO of SilentEight – an agency dedicated to protecting financial crime victims – said that this needs to be changed in order to tackle financial crime properly.
Speaking at the 2020 Dark Money Conference Thursday, Leaney acknowledged the shortcomings of technological defences against money laundering.
“For technology to work in financial crime, it needs to act as proof or something rather than as suggestion,” he said. “It needs to be completely transparent, explainable and provable. If not, it will be viewed as nothing more than ‘useful’ when it could change the game.”
Leaney advised, however, that progress had been made on technology regardless of the shortcomings, particularly in the area of using artificial intelligence to track and trace laundered transactions.
“I think AI has already gone beyond being able to join the dots like a human can,” he explained. “We see it over and over again. The AI finds potential truths in what human would deem to be falsehoods.”
“Humans get tired and not recognise flags for what they are. An AI can look at so much more data, at speed,” he said.
Global Head of Business AML at Banking Circle Livia Benisty agrees that AML technology is at least five years behind what criminals can produce.
“Take the FinCEN files leak: it almost seems like the best way of getting good AML analysis at present is to give the data to the press,” she told the conference.
“Law enforcement had been sitting on that data for too long and not acting. The press were the ones with the capacity to figure it out. Until we can get ahead of the game, I don’t think we’re getting anywhere.”
Technology in the AML world – in particular AI – is widely seen as a way to turn days of manual work into seconds of automated work, with swathes of data being read at complex levels to draw accurate conclusions about flagged transactions.
Nasdaq recently launched a new AI system which they described as able to analyse any transaction “through the lens of human-decision making.”
But experts maintain that limitations exist. Head of Financial Crime at the London-based Clearbank Marta Lia Requeijo said that progress has been catch-up in nature, merely supporting work that institutions already do and know.
“There have been huge improvements but not huge strides,” she said. “You can have the best solutions and innovations, but if you just apply what you already know, the outcome is the same.”
She also highlighted the need for top-quality data to feed into any new automated systems, advising that any technology designed to combat money-laundering is “only as good as the data you feed into it.”