Keeping pace in the financial fraud arms race

Irish Examiner

5 hours ago

With financial crime estimated to cost upwards of $3 trillion globally and continuing to grow, it can seem as if the criminals have the upper hand. But Irish fintechs are helping both financial services institutions and consumers to take on the fraudsters – and win.
'Through the deployment of technologies including digital ID checks, real time transaction monitoring, fraud awareness campaigns and artificial intelligence, Irish fintech companies are supporting the financial services systems to identify suspicious activity faster and more accurately,' says Julie Kennedy, partner in financial services risk and regulation at PwC Ireland.
Julie Kennedy of PwC Ireland.
'The Central Bank of Ireland has selected several firms for their inaugural innovation sandbox programme, the theme of which is Combatting Financial Crime,' she adds. 'It aims to provide regulatory advice and support to firms who are developing innovative methods to combat financial crime. For consumers, this means safer online banking and fewer scams and supports the development of trust within the financial services system.' Even still, it's a never-ending chase, the pace of which is accelerating. 'Criminals will exploit new technologies to defraud financial institutions (FIs) faster than they, or regulators, can keep up. With generative AI, crypto, decentralised finance, and more, financial crime can be committed in new ways and at a scale seen before,' says Daragh Tracey, director of product management at MyComplianceOffice (MCO), a compliance management software platform.
Daragh Tracey of MyComplianceOffice.
'With generative AI, fraudsters can use social security numbers stolen from a data breach and generate fake passports and fake social media profiles and then use them to open bank accounts in large volumes.' In an era of sanctions since the Russian invasion of Ukraine, criminals are using technology to create complex shell company structures distributed around the world to hide sanctioned individuals, he says, ultimately to launder money through the US financial system or other financial hubs.
'So to some financial institutions, it does feel like the bad guys are winning. FIs are burdened by slow, oftentimes manual, know your customer and anti-money-laundering (AML) systems to screen, verify and monitor customers, as well as regulatory obligations under the Banking Secrecy Act [and] the sixth AML Directive. The truth is that criminals are innovative; they'll use the latest technology to overload and bypass these checks in order to launder money and commit other financial crime.' It's why the sector must be proactive. 'Financial institutions have been in a defensive position for a long time now but those that have adapted to the new world and leveraged modern technology solutions have been able to turn the tables and get ahead of the criminals and fraudsters,' says Tracey.
'For example, where sanctioned individuals are hidden behind shell companies, we need solutions that can systematically drill down through those layers of obfuscation and identify the sanctioned person or company. If passports are being faked, FIs need systems that can automatically verify legal documents and match them against the owner. Cutting-edge software can be used to automatically risk assess not just all customers, but also all shareholders, firms and any other related business or person connected to the customer, to assess them for risk and apply scrutiny accordingly.' What doesn't work is trying to expand manual due diligence processes. 'If financial institutions are using outdated systems – sometimes even email and excel – to on-board their customers, they're going to be outpaced by criminals using the latest technologies,' he cautions.
'Using behavioural insights based on data including customer info, shareholder data, transaction activity, connections to internal employees, and more, MCO builds a network view of the risk posed by the customer that FIs can use to anticipate risks proactively rather than just reacting after the crisis.' As a fintech neobank, Revolut takes the industry-wide risk of customers being coerced by sophisticated criminals incredibly seriously and, in 2024 alone, prevented approximately €750 million in potential fraud against its customers.
Malcolm Craig of Revolut.
'The business continually enhances the app's security features, this year launching in-app calls to help customers quickly expose phone-call scams, as well as implementing real-time AI fraud detection systems, transaction limits, biometric authentication, and providing educational resources to help consumers remain informed,' explains Malcolm Craig, general manager at Revolut Bank UAB – Ireland Branch.
Indeed, Revolut's Financial Crime team represents almost a third of its global workforce. It has also developed sophisticated AI models capable of spotting and blocking potentially fraudulent transactions. Revolut, alongside many other financial institutions, deploys a number of different interventions that are designed to 'break the spell' of scammers and fraudsters, he says.
'While Revolut can't share the specifics of these interventions, so as to not provide any insight that could help criminals socially engineer their victims, the business is constantly testing a range of eye-catching warnings. Revolut urges its customers to take stock of its warnings and interventions to ensure that their money is protected,' he adds.
Current in-app security features include strict identity checks powered by machine learning, which enable Revolut to verify identity quickly and thoroughly, to prevent criminals from setting up accounts.
Every mobile device with Revolut on it is attached to a customer's biometrics. So if their phone is stolen, or their account is logged into on a new device, the same checks will apply. Customers can personalise their security settings across more than 10 features, including card controls, location-based triggers, gambling blocks, and more.
A customer's physical card must be activated in-app. That's the only place their PIN is available. 'If something looks suspicious, Revolut's machine learning models pick it up. AI algorithms can flag high-risk transactions and suspicious spending patterns immediately,' he adds.