Airwallex tackles challenge of KYC with generative AI
The cross-border payments facilitator says it has accelerated its onboarding process by 20 per cent by applying the technology to its “know your customer” compliance controls.
Payments company Airwallex is using generative AI to help comply with “know your customer”, or KYC, anti-money laundering regulations, cutting down customer onboarding times and false-positive alerts.
Government-mandated KYC rules aim to prevent financial crime by making it harder for money launderers and other criminals to finance illicit activities or move funds, and require financial institutions to apply and document identification procedures for their customers.
Co-founder and CTO Jacob Dai said AI helped tackle the “challenge” of compliance with stringent KYC rules.
“KYC is that vital first step in our relationship with a customer.”
“We use a combination of intelligent technologies and human expertise to establish that new customers are who they say they are, that they’re being truthful about the nature of their business, and that their products or services are legal and in line with Airwallex’s acceptable use policies.”
Airwallex was founded by Mr Dai and three other friends from the University of Melbourne in 2015 to facilitate cross-border payments for businesses and has since grown its customer base to over 100,000 worldwide.
Previously, it used rules-based analytics and natural language processing to scan new customers’ websites to flag suspicious applications but encountered a “significant volume” of false-positive alerts.
It said adding a generative AI co-pilot that used large language models reduced false positives by 50 per cent when scanning for “red flags” during the application process and sped up the KYC onboarding process for customers by 20 per cent.
Mr Dai said generative AI had enabled customers to be onboarded within minutes while enhancing the platform's fraud detection and prevention capabilities.
“With generative AI, we’re honing our KYC tools to be much more accurate and context-aware,” he said. “These enhancements translate directly to our customers.”
“Generative AI helps Airwallex’s models to be much more sensitive to the context and meaning of keywords. When scanning a customer’s website, the model can better distinguish between a retailer selling a ‘military-style’ jacket and a merchant selling prohibited military goods, or a ‘champagne-coloured’ dress and actual champagne, or smartphone accessories versus actual smartphones being sold illegitimately.”
Airwallex said its KYC AI co-pilot would be the first of several AI-based tools it would be rolling out, and that customers could expect product features and operational enhancements next year.
“There are many applications to explore, and we’re just getting started,” Mr Dai said.
About the author
