On Wednesday 12 February 2020, Vocalink participated in the UK Finance Economic Crime Congress in London, UK. Our own David Rich, Executive Vice President, and Nicole Tuckwell, Director of Product for financial crime solutions, shared insights and solutions in the fraud prevention stream.
Reprising our roaming reports from Sibos 2019, read our summary of the day’s discussions reported under Chatham House rules.
Fighting fraud: why the use of artificial intelligence is key to uncovering new, increasingly sophisticated behaviours
According to data published by UK Finance, economic crime costs the UK nearly £7 billion a year. It affects all parts of society, including businesses and consumers. So what are we doing about it? David Rich, who heads up the Vocalink financial crime solutions team, opened the stream with a presentation on how artificial intelligence (AI) is enabling the fight against financial crime.
Our solutions, he noted, leverage high-volume and real-time network level data. The thoughtful application of machine-learning algorithms enable us to detect patterns that aren’t visible to humans, and across larger data sets at much faster speeds. As the algorithms ‘learn’, they become more accurate.
In 2019, we partnered with NatWest to help prevent business payment fraud. We analysed the behavioral signatures left behind every time a business pays an invoice and signatures of known historical frauds, and trained machine learning algorithms to identify and flag suspected incidents of fraud before they occur. As of September 2019, the solution had prevented losses of £14.5 million with an in-flight detection rate of 90 percent. A fraud feedback loop continues to evolve the machine learning models.
Employing AI in the fight against organised crime is now necessary to level the playing field: We’re seeing evidence of international money laundering networks using artificial intelligence to open accounts and automate transfers as a human would — small, irregular transactions that appear more ‘normal’.
Criminals are increasingly smart — but we’re smarter. We’re leveraging big data insights to hone our algorithms and exploit criminals’ mistakes. And we’re increasingly looking at indicators of suspicious activity beyond the transaction itself: inconsistencies in IP addresses and device IDs used to access mule accounts.
While AI is the way forward, humans remain an essential part of the equation. “We see technology as an enabler: We use machines to do the ordinary and people to do the extraordinary.” As one delegate observed, AI helps to reduce the “soul-destroying” and inefficient task of investigating false positives, freeing up humans to do more valuable work. Together, human and artificial intelligence will win the fight against organised crime.
Don't be fooled: Money Mules
After lunch, attendees were asked what more the industry could be doing in the fight against money mules and their recruitment. Our own Nicole Tuckwell was joined by representatives from CIFAS, LexisNexis, Barclays Bank, and the Cyber Crime Unit of the Metropolitan Police.
One new and frightening trend, the panel observed, is the use of temporary mules: where people’s accounts are taken over for just a short period of time (sometimes without them knowing), or where criminals groom and incentivise individuals to ‘borrow’ their bank accounts to move ill-gotten funds. These transactions don’t stand out in account activity because the amounts are generally small.
Mule herders (those controlling networks of mule accounts) recruit prolifically using social media and instant messaging channels, luring in young people by the idea of easy money. Many young people seem unaware of the risks they’re running and the consequences it can have not only for them, but for society as a whole. According to figures from fraud prevention service Cifas, as many as 10,686 under 21s were caught up in the crime last year – a rise of 26% on the previous year.
People and businesses must be better educated on the risks and how to protect themselves from fraud, notably authorised push payment scams. Research into social conditioning methods employed by organised criminals will help to inform education initiatives. In parallel, the panel agreed the efficacy of any solution to combat money laundering and money mules depends on industry collaboration; overcoming barriers to sharing data, in particular in real-time.
In 2018, Vocalink partnered with Pay.UK to develop the anti-money laundering Mule Insights Tactical Solution — a world-first industry-level solution to trace financial crime and alert on suspect mule accounts across the UK’s payment networks. We continue to enhance the solution based on learnings and insights a year since launch and stay ahead of the latest criminal trends; our intelligent algorithms, tested and trained on high volumes of network-level payment data, are available to participating financial institutions and schemes around the world.
Highlights from elsewhere at the event: