Podcast: Play in new window | Download
Artificial intelligence (AI) and biometrics are revolutionizing regulatory compliance in fintechs and banks by providing more accurate and efficient methods of identifying and preventing fraudulent activity, as well as streamlining compliance processes.
Traditionally, compliance has been a tedious and time-consuming process, requiring manual checks and reviews of transactions and documents. But with the help of AI and biometrics, compliance is becoming a lot more efficient and effective. In a recent PaymentsJournal podcast, Micheal Sheehy, Chief Compliance Officer at Payoneer, and Marco Salazar, Director of Technology and Infrastructure at Javelin Strategy & Research, discussed the future of meeting compliance challenges.
The Future of Compliance Challenges
The biggest challenge for fintechs in compliance is the cost of implementing Know Your Customer (KYC), a process fintechs use to verify the identity of their clients and assess their potential risks for money laundering or financing terrorism. Fintechs may need to go through a KYC process when onboarding new customers, setting up new accounts, or conducting certain financial transactions. This typically involves collecting and verifying personal and financial information, such as name, address, government identification, and employment status. Fintechs may also need to monitor their customers’ activity over time to ensure ongoing compliance with KYC requirements.
“Especially when you want to be global and operate in multiple jurisdictions, you know, the different KYC nuances can be costly,” explained Sheehy. “The repercussions of not having an adequate KYC program or adequately funded compliance programs are significant. [That] there are $10 billion in just KYC fines last year globally just shows you how serious regulators are taking KYC.” Furthermore, different countries are developing different regulations so staying on top of everything is a challenge.
“Criminals are always trying … to find loopholes in the system,” Sheehy said. “So [compliance] is about being proactive. This involves having processes and procedures in place to analyze the trends that you’re seeing not only in your own transactions, but also at a more macro level within the environment that you operate in.”
As a company that interfaces with regulators and fintechs looking to meet those regulations, Payoneer acts as a steward of the global economy and makes the complex world of regulatory compliance simpler. “The complexities outlined by Micheal drive this desire for simplification, which will require an iterative process to attempt to get there,” Salazar said.
To meet different stringencies of KYC regulation around the world, many companies use the approach of just trying to meet the strictest requirements. But this can backfire for companies based in heavily regulated countries, such as Singapore, seeking to grow globally. For such companies, “when you’re dealing with customers in the U.S., where the KYC requirements aren’t as stringent in the regulations, you’re putting yourself at a competitive disadvantage compared to your other peers that may not be operating globally,” Sheehy said. Complying with local regulations is challenging even for the biggest multinational companies. “Apple and Google are trying to scale globally but are restricted by local legal mandates,” Salazar said. “They’ve run into regulatory issues where they have to decide whether to [incur] fines or scrap complete products.”
The Role of AI in Payments Management
One way AI is improving compliance in fintechs and banks is through the use of machine learning algorithms. These algorithms can analyze vast amounts of data, identify patterns and trends, and make predictions about future events. This enables banks and fintechs to identify and prevent fraudulent activity before it occurs, rather than reacting after the fact.
“Historically, compliance was, you know, detect and report, detect and report. Now we’re moving into effective prevention and also more real-time reporting,” Sheehy said. “Machine learning and AI really allows you to operate in more of a real-time environment versus a traditional rules-based environment. The traditional model involved using rules such as if A happens, do B, or if C happens, do D. In contrast, machine learning will enable you to enact preventative measures and have more insight into how your customers transact. And it also enables you to operate in a more real-time manner.”
For example, Sheehy described how Payoneer used AI and machine learning to model merchant behavioral patterns in a certain jurisdiction selling certain goods. “Is a merchant new to the market? Or is it an established merchant and [has] been operating for 10 years? You’re not going to treat them the same,” Sheehy said. “Someone who’s growing and starting a business will have smaller payments that ramp up over time. A more established customer that will have large volumes that peak throughout seasonal periods.”
AI models can help fintechs segment their merchants by type and predict what will happen in the future. “If somebody receives a large payment, your model could say, well, I think that x is going to happen. This could trigger a request for additional KYC verification or pause that customer’s activity.
With AI, machine learning models can be tailored to specific countries or markets. “With the emergence of technology and new platforms, we’ve had this acceleration of data governance standards, even though they’re still very disparate across regions,” Salazar said. “We’re starting to see the ability for these models to really learn and … drive impact within those regions, which makes a big difference.”
Biometrics and Compliance
Alongside AI, biometrics is also making waves in the compliance world by using physical characteristics for identification and authentication. This allows customers to easily access their accounts by simply looking into a camera, eliminating the need for passwords or other forms of authentication. Banks are also using voice recognition software to verify the identity of customers over the phone, as well as fingerprint scanners to ensure secure access to accounts. It’s a lot harder to impersonate someone else’s facial features or voice or fingerprint than it is to guess their password.
“Everybody uses biometrics, when they unlock their phone, when they use Apple Pay, when they use a fingerprint on something. It’s already kind of a standard,” said Sheehy. “I think biometrics is tied significantly with digital identities, which I’ll go into in a second. After the Equifax data breach, COVID unemployment scams, and PPP loan scams using stolen identities, it really became obvious that the only way to prevent this fraud is a live biometric check. Tying this together with digital identities is super important. By leveraging a government database to pull someone’s digital identity and cross-checking it with a biometric test, you can tie the two of them together.”
Globally, digital identities and biometrics are much more advanced in Africa and Asia, with Europe and the U.S. lagging somewhat. But Sheehy claimed that biometrics will be the standard globally within the next two years. “Singapore and Malaysia have actually mandated biometrics in their KYC. They’re telling the financial institutions in those markets, if your customers are not in front of you when you’re selling financial products, you need to have a liveness and KYC check. They go so far as to claim that they will not accept identity theft as a typology within their economy anymore.”
Artificial intelligence and biometrics are more than just cool gadgets — they’re improving the compliance function in fintechs and banks in a big way, helping keep our money and assets safe and secure. Biometrics are still not perfect, “but it’s a significant change from five years ago, where people were just taking pictures of their IDs and uploading them and applying for mortgages and things like that,” Sheehy said.
In the United States, looking forward, for biometrics to have wide-scale adoption, it requires standardization and government regulation around data. “Right now, regulation of biometrics is at the state level. We need more of a federal mandate, which I believe is coming. Until then, it’s kind of the Wild Wild West.” Part of this regulation could be in the Consumer Data Privacy Act that is currently being debated in Congress.
As various KYC regulations change throughout the world, Sheehy is optimistic that Payoneer can be part of the solution in making payments more secure while complying with regulations and innovating in machine learning and biometrics. The future certainly seems bright for companies that can help simplify international regulatory complexity while making better use of customer and business data.