Sibos 2024: A delicate balancing act – addressing the ‘speed vs security’ dilemma in cross-border payments
A major trend today is that traditional offline B2B trade is increasingly moving online. This shift has resulted in a boom in the number of B2B marketplaces.
In today’s world, customers are constantly demanding faster and more convenient cross-border payments
The explosive growth of online B2B trade has been a key driver of cross-border payments in recent years. But this has also presented a myriad of challenges.
In the opinion of David Messenger, a long-time participant in the global fintech industry: “The benefit of this is that, when you digitise [trade], you inherently are more data-rich, and more data-enabled than in an offline analogous business model.”
“B2B trade definitely is higher-risk than B2C e-commerce marketplaces,” says Messenger, who is CEO of Global Businesses at PingPong, an international cross-border payment service provider.
As he sees it, part of that risk stems from more stringent anti-money laundering (AML) requirements for the financial institutions and fintechs that serve B2B platforms.
With massive amounts of trade data now digitised and multitudes of new players introduced into the global payment ecosystem, Messenger believes financial service providers need to have the right tools in place to manage the higher risks of money laundering and fraud.
“This is where you need to be deploying new machine learning and AI tools and look for ways to standardise data, and share data between institutions,” he states.
Messenger is not alone in advocating a bigger role for cutting-edge technologies in mitigating ever-evolving risks in the B2B sector.
Lyu Weiyan, Co-CEO of LianLian Global, a Chinese platform for cross-border payments, has had a front-row seat to witness the profound changes in China’s cross-border B2B e-commerce landscape over the past few years.
The adoption of AI in particular can deliver a boost in the fight against fraud. Typical use cases include leveraging it to analyse transaction models and conduct risk scoring to identify abnormal activities “within seconds of a transaction,” she says.
Lyu has noted a significant shift in cross-border B2B e-commerce in that customers no longer operate in just one or a few countries, with the aim instead to forge a global payment and settlement network.
A trade-off
Amid this fast expansion, customers are constantly demanding faster and more convenient cross-border payments. This expectation sometimes clashes with ever-tightening AML regulations and protocols in almost every jurisdiction.
“One thing I would highlight is that in an era of heightened risk, there is always a trade-off between managing risk and fulfilling the customer’s expectations and delivering the service that we are here to do, just to facilitate trade,” Messenger says.
He explains the financial services industry is permanently looking to harmonise “speed and effectiveness” from a customer’s point of view and “managing the risk” from the regulatory perspective. “That’s probably one of the biggest tension points for all financial institutions to be managing,” he adds.
Financial institutions are also performing a delicate balancing act between rising compliance standards and stricter legislations around user data security and privacy.
“The key to managing AML and fraud risks in cross-border transactions really comes down to data,” says Messenger. “It comes down to having consistent data and protocol across different regions, because the more data you have, the better able you are to detect and manage risks and fraud.”
The role of PingPong, and that of many similar companies, puts them in an advantageous position. Having access to information from both sides of the transaction allows them to double the data, enhancing their ability to detect fraud and identify broader patterns within the settlement network.
But tensions are likely to persist, as many tend to believe both being faster and being safer cannot be reconciled in cross-border payments. Recently, there have been attempts to address the risk management conundrum from a fresh perspective.
Traditionally, AML procedures centre on the sender instead of the recipient. But according to Rod Francis, a partner in Oliver Wyman’s Financial Services Practice, some industry practitioners are calling for more attention to be directed toward the latter.
This would require the recipient’s bank or fintech partner to undertake risk scoring around the client before deciding to release funds either in whole or in part, says Francis, who is a Hong Kong-based expert specialising in anti-financial crime and regulatory compliance.
Otherwise, under the unrelenting pressure to be quicker in processing cross-border payments, conducting risk management would become extremely challenging in all classes of financial crime, including fraud, he adds.
The one-million-dollar question
Despite the innovative proposal, Francis says that the “faster vs safer” dilemma is the “one-million-dollar” question to tackle, with no “silver bullet” in sight.
However, Lyu of LianLian believes that speed and security don’t necessarily have to be at odds. While financial service providers worldwide are struggling to strike the right balance between “faster” and “safer,” many, in her view, have overlooked the role of technology in reconciling these two goals.
Aided with advanced technology, “speed and security surely can become complimentary,” she claims. “We just need to improve our risk control technology while enhancing our efficiency.”
As positive as this outlook is, the fast ascent of emerging technologies also brings with it its own challenges. Powerful tools like generative AI automate lots of work, making life easier for financial professionals, but they also complicate risk management in cross-border payments.
Recent years have seen a spike in cases where fraudsters have tried to circumvent risk control systems using AI-generated synthetic identities and deepfakes. “So it’s a double-edged sword,” Messenger says.
With both good and bad actors in possession of advanced technology such as generative AI, it becomes tricky for financial institutions, banks and fintechs to ensure they stay one step ahead.
“Previously, the pace at which new frauds happen and can be detected is slower. With the rise of these powerful new tools, that pace is accelerating,” Messenger says. “Therefore, the approach on the side of financial institutions needs to have that sense of urgency in terms of developing tools and recognising that new trends and new fraud risks will appear very quickly and require faster response.”
Francis concurs. He acknowledges the vast potential for AI misuse, which is already occurring, but emphasises its role in advancing risk management. In some areas of financial innovation, this is being robustly implemented and explored.
Possible applications are many. For example, when submitting a suspicious activity report, AI can analyse investigation data to draft the report, covering 80-90% of the necessary details. This helps articulate the grounds for suspicion, clarifying why the activity raises concerns in relation to fraud or money laundering.
But he warns that when using generative AI, it’s essential to conduct thorough reviews and testing to ensure desired outcomes and to comply with mounting regulatory scrutiny.
Comparing AI’s use from five years ago to now, Lyu has noted many differences. Initially, AI was mainly used in generative applications like knowledge bases and Q&A systems. Now, it is integrated into operational workflows, such as automated screening during technical development reviews.
“However promising they are, at the end of the day, new technologies must be assessed on their risks before they can be applied for greater efficiency, compliance and safety,” she says.
Thank you for your analysis on the challenges and opportunities facing financial institutions in cross-border payments, particularly in the context of anti-money laundering (AML) regulations and fraud prevention, is insightful. While today these issues may appear novel and fraught with risk, it is crucial to recognize that the future will likely center on how international trade can continue to thrive alongside these evolving barriers.
As technology and practices improve, compliance with AML regulations will become more achievable. However, it is essential that governments interact and exert oversight over the cross-border B2B market to advance in this path. Such progress not only benefits states but also facilitates smoother operations for business-to-business clients and suppliers.
In recent years, some governments have taken bold steps to embrace cyber-commerce and manage their digital assets effectively like Estonia’s Digital Economy and Singapore’s Smart Nation Initiative for example. This type of forward-thinking approach needs to be adopted by key global powers to ensure that we can achieve comprehensive, effective oversight and technological integration on an global scale.
Collaboration between regulatory bodies and advancements in technology, particularly artificial intelligence (AI), can streamline fraud detection, making it faster and more efficient. AI, fed with extensive metadata and real-time data sharing, can make the processes of detection and prevention almost fully automatic. This is critical given that fraud has reached unprecedented levels, with tactics such as AI-generated synthetic identities becoming more common.
In conclusion, while today’s challenges in combating fraud and complying with AML regulations might seem formidable, they also present an opportunity. By fostering global cooperation and adopting a proactive mindset towards technological and regulatory integration, we can create a landscape that safeguards international trade while promoting trust and security for all stakeholders.