Interview: Neo4j Global Head of Finserv Michael Down on the $442bn fraud problem banks can’t see

Some $442bn has been mislaid globally to fiscal fraud successful the past twelvemonth and arsenic banks astir the satellite accelerate their displacement to data-driven fintech models, the complexity of these systems is creating caller unsighted spots for fraud.

As Michael Down tells RBI, fraud present moves done coordinated networks of mule accounts, synthetic identities, and transactions, alternatively than successful isolation.

As AI is helping to summation the speed, standard and sophistication of attacks, the real-world interaction is 1 successful which fiscal firms aren't presently acceptable up to link the dots wrong these invisible networks. And helium explains however graph quality goes beyond detection, to alteration what's operationally imaginable for the teams down the scenes and is palmy successful uncovering suspicious patterns that accepted systems miss.

RBI: In your opinion, however is AI expanding the speed, standard and sophistication of attacks successful the FS sector?

Michael Down, Global Head of Financial Services astatine Neo4j:

AI has fundamentally changed the economics of fraud. What erstwhile required coordinated quality effort – creating fake identities, crafting convincing phishing messages, and manipulating documents – tin present beryllium automated and deployed astatine standard successful minutes with AI.

We're seeing this play retired crossed aggregate fronts: deepfakes are utilized to bypass individuality verification systems done cloned voices, manipulated video and facial imagery. Separately, AI-generated synthetic identities harvester fabricated idiosyncratic accusation and manufactured integer footprints to make accounts that look wholly legitimate, portion AI-powered phishing campaigns continuously trial and refine the approaches astir apt to succeed. The effect is fraud that is much convincing, much coordinated, and much blase than thing the manufacture has seen before.

The situation for fiscal institutions is that bequest detection systems were built for a antithetic menace exemplary - 1 wherever anomalies appeared individually and could beryllium caught done rule-based controls. Today's AI-enabled fraud deliberately spreads enactment crossed accounts, devices, and institutions to debar triggering those rules. Traditional systems, therefore, request to adapt.

RBI: You person referenced enactment your steadfast did with BNP Paribas Personal Finance to trim fraud by 20%...can you springiness immoderate accusation connected however the solution achieved this?

Down: We helped BNP Paribas Personal Finance fundamentally displacement from treating fraud arsenic an isolated lawsuit to helping them place connected fraud networks.