Mon, 24 Mar 2025
Fraud prevention in financial services is a never-ending arms race. As business expand their digital presence, financial product offerings, and payment methods become more seamless, fraudsters find increasingly sophisticated ways to exploit vulnerabilities. Adyen, a global payments platform and licensed bank, stands at the forefront of this battle.
In a conversation with Miguel Rivera, Head of Risk Products at Adyen, we explore the evolving fraud landscape, the unique challenges fraud prevention teams face, and how Adyen’s vast dataset provides a competitive edge in keeping merchants and customers safe.
Fraud is not what it used to be. Traditionally, financial crime was dominated by third-party fraud—where stolen credentials were used to conduct unauthorised transactions. However, Miguel highlights a critical shift: fraudsters are now increasingly engaging in first-party fraud, a more insidious form of deception where legitimate customers manipulate systems for financial gain.
One of the fastest-growing trends in first-party fraud is policy abuse, where customers take advantage of return policies, promotional offers, or refunds. Miguel illustrated this with a real-world example:
"A customer buys a luxurious bag, flaunts it on Instagram for the weekend, and then returns it, claiming a defect. From a transactional perspective, everything appears legitimate, but this is a form of fraud that merchants are struggling with."
Mastercard reports estimate the annual cost of first-party fraud at approximately $50 billion, and merchants are bearing the brunt of these losses. Since this type of fraud does not involve unauthorised access or chargebacks, it is much harder to detect using conventional fraud detection methods.
Another area of concern is Authorised Push Payment (APP) fraud, where scammers trick individuals into sending money voluntarily under false pretences. This type of fraud has become more sophisticated with AI-generated voices and deepfake technology.
In the UK, regulators have introduced liability-sharing between banks and victims, creating pressure on financial institutions to step up their prevention efforts. While Miguel acknowledges that customers bear some responsibility, he argues that the industry has a duty to educate and protect users:
"We can’t just blame the customer for being tricked. Just as a parent educates a child, we have a responsibility to guide our customers and implement safeguards to prevent these scams."
Artificial intelligence has supercharged fraud detection, but it has also empowered fraudsters. AI now allows criminals to create highly convincing synthetic identities that can pass traditional Know Your Customer (KYC) checks.
"Fraudsters are not regulated, so they can move faster than us," Miguel explained. "They can create synthetic identities with real-looking documents, blending legitimate and fake information to bypass security checks."
This evolution means fraud prevention systems must adapt dynamically, leveraging AI and machine learning to stay ahead.
As both a payment processor and a licensed bank, Adyen is in a unique position—its extensive dataset allows it to detect fraud patterns across multiple merchants, industries and geographies, giving it an edge over traditional banks that only see their own customers’ activity.
"If a customer engages in fraudulent activity with one merchant, they are likely doing it elsewhere. Individual merchants can only see part of the picture, but we see the full pattern."
However, Cross-merchant visibility enables Adyen to proactively detect and mitigate fraudulent behaviour while adhering to strict regulatory and compliance standards. As a bank and regulated financial institution, Adyen applies a cautious approach in selecting use cases that leverage their data intelligence, ensuring both effectiveness and data protection.
Balancing security with user experience is a key challenge in fraud prevention. Overly strict measures can create unnecessary friction, frustrating genuine customers, while lenient measures expose businesses to risk.
Adyen addresses this challenge by using risk-based authentication, which adapts security measures based on customer behaviour.
"If we recognise legitimate users, we can reduce friction, while unusual behaviour might trigger extra scrutiny. For example, a sudden high-value purchase in Latin America by a European shopper might seem suspicious and trigger additional scrutiny. However, it could happen as well that based on our vast data that allows us to assess historical patterns, we might identify that while initially suspicious at a first glance, the transaction from the example is indeed legitimate and should be approved"
This adaptive approach enhances security while ensuring seamless transactions for trusted customers.
Adyen employs machine learning models to detect fraud in real time. However, Miguel acknowledges that AI is not foolproof. While automated systems are excellent at spotting anomalies, at this point in time, human oversight remains essential, particularly in areas like Anti-Money Laundering (AML).
"For AML alerts, we use AI to flag potential risks, but human analysts conduct manual reviews. Fraud prevention isn’t just about blocking bad actors—it’s also about ensuring that good customers can transact freely."
This hybrid approach allows Adyen to strike the right balance between automation and human expertise.
While Adyen’s fraud prevention capabilities are advanced, Miguel believes that collaboration across the industry is necessary to tackle fraud effectively. However, regulatory barriers often prevent financial institutions from sharing data, creating gaps that fraudsters exploit.
"If we could share fraud intelligence at an industry level, it would be a game-changer. But regulations limit how we can use data, even when it’s in the best interest of fraud prevention."
Miguel argues that Regulators need to adapt to the fast-moving fraud landscape and work more closely with industry leaders to create frameworks that support innovation without compromising compliance.
Looking ahead, Miguel predicts that fraudsters will continue to refine their tactics, taking advantage of AI, regulatory blind spots, and global inconsistencies in fraud prevention.
He points to organised fraud syndicates that operate on an industrial scale, such as fraud factories in Myanmar, where human traffickers force victims to conduct online scams.
"This isn’t just small-scale fraud. These are highly organised operations, complete with managers, targets, and training. Fraud is becoming a business in itself."
This underscores the need for financial institutions to think globally and use AI and data-sharing to combat these sophisticated threats.
For Adyen, fraud prevention is not just about stopping losses—it’s about creating a safer and more efficient payment ecosystem. By leveraging its extensive dataset, AI-powered risk detection, and industry-leading fraud prevention tools, Adyen helps merchants increase authorisation rates by up to 6% while keeping fraud levels stable.
"Fraud prevention isn’t just about blocking bad actors—it’s also about ensuring good transactions go through seamlessly," Miguel emphasised.
As fraud continues to evolve, companies that prioritise data-driven, AI-enhanced, and customer-centric fraud prevention will gain a competitive advantage—not only by reducing losses but by building trust in the digital economy.
You have the opportunity to catch up with Miguel in person at our flagship conference in Brussels on May 22, where he will be sharing his insights during a panel on "The Tech Advantage: Balancing Risk, Fraud, and Cost Management Effectively" with fellow industry experts.