Fighting financial crime is already an inherently difficult task because it requires finding deliberately obscured patterns and connections not just within, but across several massive sets of data. And it’s not going to get any easier. The whole financial landscape is changing, and while fintech gets a lot of buzz, it’s only one of several factors adding more dimensions to anti-fraud efforts:
Increasing consumer demands…
People want instant access to their money, especially from mobile devices. Payment services like Zelle, Venmo, and Cash App have risen to meet this need while providing minimum friction, leveraging the convenience of nearly ubiquitous smartphones, and in the case of Zelle, widespread integration into the mobile banking apps of major banks like Wells Fargo and Bank of America.
….met by an increasing number of options
However, those apps aren’t the only technology on mobile shaking things up for finance. The almost standard inclusion of near field communications (NFC) in smartphones has spurred a digital analogue to an analog financial mainstay: the digital wallet. Pushed by big tech companies like Apple, Google, and Samsung, these apps essentially transform a mobile device into a debit card, linked to a digital banking account separate from the user’s main checking or savings. The short-distance nature of NFC, use of strong encryption, and the elimination of the need to pull out a card to tap at the point of sale make digital wallets a secure and convenient way to pay merchants.
The innovations in banking don’t stop with hardware either. Bitcoin and other cryptocurrencies continue to creep further into the mainstream, bringing with it wider legitimacy of its two core ideas: the distributed ledger (the blockchain), and anonymous transactions.
The convergence of all these innovations – and their accompanying legitimate transaction patterns – increases the data modeling challenge that compliance analysts must tackle as they attempt to sift through the data, looking for abnormal behavior which can indicate fraud.
Meanwhile, the criminals are increasing their sophistication, tweaking their fraud, money laundering, and other financial crime tactics as they learn from past successes and failures. From adopting Bitcoin to hide their tracks to using Venmo to scam victims out of thousands of dollars, the bad guys are finding lots of cover in this new financial technology landscape.
Anti-Fraud Systems Must Evolve to Meet This New Reality
What does all this mean for financial crime management? Rapidly increasing workload and compliance team headcount, for one. It also means greater chances of financial crime going undetected, higher volumes of false positives in traditional anti-fraud systems, and more time spent on low-risk incidents.
The old tools are failing to keep up with these new trends and demands. We need a new approach to the whole financial crime management workflow: detection, investigation, and reporting. Upstream from that workflow, however, is the challenge of corralling the firehose of data coming from many sources, in multiple formats, and at varying levels of completeness. These data and analytics problems must be solved.
Fortunately, a new type of anti-fraud system has emerged which combines data, analytics, and process automation in a way that improves detection while reducing costs. In these solutions, machine learning extracts actionable intelligence from the raw data, with advanced analytics providing a window into that intelligence at every step of the process. In this way, human judgment is augmented by intelligent automation, resulting in quicker resolution and action.
Improved operational efficiencies inside compliance teams, lower compliance risk, and much less time spent on false positives are additional benefits of autonomous financial crime management. This smart combination of human experience with the inherent adaptability and self-tuning detection models resulting from AI is the approach needed to meet the scale and complexities today’s banking and finance world.
The future of financial crime management isn’t fully autonomous, just as its past wasn’t without its software tools. The greatest value will continue to lie in an optimal mix of the two, a combination which makes this inherently difficult task much more manageable.
Image by iAmMrRob pixabay