Next Generation Identity Matching: How to Reduce Regulatory and Operational Risk
Jordan Lo, Senior Project Manager at Swiss RegTech leader IMTF, is passionate about identity matching. He knows that the crux of dealing with many regulatory challenges is identity matching, which if properly thought through and executed, can eliminate a considerable amount of inherent regulatory and operational risks. He addressed the audience at the Hubbis Compliance in Asian Wealth Management Forum to explain IMTF’s state-of-the-art solutions , and more specifically IMTF’s cutting edge identity matching technology.
Lo began his talk by acknowledging that technology is not a cure-all. “It is not the whole solution,” he said, “but it is one of the key pillars that we must bring “in the fight”. And the ”fight”, by the way, is not with the regulators, but with the money launderers, the terrorist financiers and the cyber-criminals.”
He referred to the 10-year challenge - a current social media phenomenon of people comparing pictures of themselves today and a decade ago - and explained that while many celebrities look similar to how they looked ten years ago, the world of technology and RegTech certainly does not. “A decade ago, there was little talk about regulation, little talk about the use of technologies to fight financial crime. But our world has changed dramatically. The controversial social phenomenon is also a huge topic in the AI world (facial recognition training) and how it may be used or misused. With that he dived back into technology and AI from an IMTF perspective.”
IMTF’s deep experience
Founded in 1987, IMTF is an international software and application integration company headquartered in Switzerland with offices worldwide, such as in Singapore, to meet the demand of their growing client base in Asia. The company offers innovative software solutions which enable clients to increase efficiency, achieving significant cost reductions with assured compliance.
IMTF builds RegTech software solutions to improve operational efficiencies, manage regulations and offer new services like engaging with customers across all channels. The comprehensive offering, competencies and tools focus on: onboarding and client lifecycle management; AML, tax and market monitoring compliance; smart screening and open source investigation, custom-built front-ends for all types of channels and devices; technology and process consulting; professional services for the successful implementation of complex projects; secure document management; adaptive case management and collaboration.
“IMTF has acquired experience in compliance and FI challenges over the past 3 decades," Lo reported. "We understand our clients, and are dedicated to solving their issues in fields such as transaction screening, fraud detection, and identity matching. In today’s presentation, I am focusing on the challenges of identity matching, as it is so central and crucial to our clients. It is all about reducing false positives, reducing false negatives and thereby mitigating both operational risk as well as regulatory risk."
Cutting down the false leads
He explained that, in simple terms, the fewer false positives, there is less noise for FI operations teams to investigating. "We know that today so many FIs are operating in a zone of frustration, and we also know that without efficient, accurate identity matching they risk all their other capabilities not performing as expected, meaning that there is a negative impact throughout, for example on trading, compliance, transaction monitoring, fraud detection, and so forth."
Lo noted that the expectations of the regulators have changed drastically. “Words and phrases such as robust detection escalation capabilities, event triggers, risk assessment capabilities, parameters, thresholds and scenario setting all tell us that the demands from the regulators are rising all the time and technology is the key enabler to meet these expectations. There is an expectation that FIs not only leverage the use of technology but also in ways that meet their expectations in terms of functional, operational capabilities and customizable to sure sustainability in dealing with future threat evolutions.”
Much more can be achieved
He highlighted some of the methods for identity matching today, such as manual searches, real-time screening and matching software, distance measures and others. Lo then covered some of the sophisticated features of the IMTF identity matching technology, such as unmasking the same identity even when alternate names are used, or even other languages. He explained that when electronic name screening was introduced twenty years ago, only a few hundred names were involved. Now, millions of names are screened worldwide all the time and against hundreds of lists containing millions of names. The cultural context of names has now to be taken into account and identity matching solutions have to make accurate distinctions between names from different cultures, aka in different languages, including transliterations. Finally, for Enhanced Due Diligence, the searchers’ intent has to be supported by allowing to combine names with risk categories to faster identify relevant content.
“For example,” he reported, “we can do so many things related to the Chinese language, in order to detect matches and reduce false positives. Artificial intelligence (AI) and machine learning are at the core of our solutions as your name matching capabilities must be top notch and to achieve that we offer smart and tactical matching, phonetic matching, ethnicity awareness, smart semantic matching and other protocols that help IMTF's FI clients achieve their optimal outcomes.”
IMTF forging the way
Lo reported that IMTF has achieved with its NextGen name screening to reduce false positives by 25-75% and cut down the review time so as to decrease operational and regulatory costs, benchmarked against other typical algorithms on offer in the RegTech universe. “We are very excited about this, and we want you all to know about it,” he stated.
Lo then went into more detail about event triggered matching and alert generation. “We know that current solutions are not adequate,” he explained, “so we have developing technologies in AI to recognise if a target had triggered an alert before so that the system should intelligently remember when the alert was triggered and in what context and why."
Then, if and when something material changes, new triggers arise but also refer back to past alerts. “We are trying to connect technology and human skills,” Lo explained, “so that they go hand in hand to help the users in a workflow collaboration.”
Taking the supervised and unsupervised routes
Demystifying the new machine learning techniques, Lo then introduced some of the new advances IMTF is pioneering. “Supervised machine learning is about training the logic,” he noted. “By introducing your own rules, you can achieve supervised rule engine building and machine learning whereby the technology builds the algorithms itself. We can also do unsupervised machine learning, which with the right infrastructure in place and data of sufficient standards we can detect certain patterns that uncover interactions and relationships between parties and transactions.”
Lo concluded by advising the audience not to fear technology, but to embrace it. “We live in a complex world, “he observed, “and we all know how many parties there are involved in a bank or other financial institution. At IMTF, identity matching is a core capability in our modular RegTech platform and enables all other compliance modules. My final message is that with the right technology, it is most certainly possible to derive information and detect patterns that will allow our FI clients to fulfil their internal and external obligations.”
More from Jordan Lo, IMTF
Latest Articles