Payments security has reached a crossroads in terms of speed, connectivity, and global reach. For years, financial institutions strengthened individual controls, authentication, transaction monitoring, sanctions screening, and fraud detection. Those controls remain essential, but they are no longer enough on their own. Fraud never appears at a single point in the transaction. It develops across a customer journey that begins long before a payment is initiated and continues well after it has been processed.
At Bottomline’s Sibos 2026 session, Securing Payments End to End: AI, Trust and Security in a Real-Time World, I heard that shift expressed clearly. Moderated by Jessica Cheney, VP Banking Solutions and Growth at Bottomline, the discussion brought together Giorgio Andreoli, Director General of the European Payments Council, and Genevieve Gimbert, Lead Client Principal and Fraud Solutions Offering Leader at PwC US.
What emerged from the discussion was a growing recognition that fraud is no longer just a payments problem. It is an interconnected challenge that spans identity, customer onboarding, authentication, transaction processing, and post-payment monitoring. In my many in-depth conversations with banks and financial institutions around the world, this is one of the most common themes. Fraud teams are often focused on the payment itself, while attackers exploit vulnerabilities much earlier in the customer journey. By the time a suspicious transaction appears, the compromise may already have happened during onboarding, authentication or account access. As payments become more real time, those connections matter more than ever.
Moving Fraud Prevention Upstream
One of the more interesting insights from the session came from comparing audience views with Bottomline’s own research.
When attendees were asked where they believed the biggest fraud-related gap exists, pre-payment validation topped the list. Meanwhile, Bottomline’s Payments Intelligence Gap research, based on feedback from more than 300 payments professionals, found that mid-transaction fraud detection remains the leading concern as cited by 44% of respondents, followed by pre-payment validation at 28%, and end-to-end visibility at 19%.
Taken together, the results suggest that while financial institutions are investing heavily in fraud prevention, many continue to struggle with visibility across the wider payment journey.
What I find particularly interesting is that many banks and non-banking financial institutions already have effective fraud controls in place. The challenge is often that those controls operate independently of one another. As fraud becomes more sophisticated, institutions need to think less about individual checkpoints and more about how intelligence flows across the entire customer and payment journey.
“Fraud is rapidly evolving, requiring financial institutions to protect transactions across the entire payments lifecycle, from initiation and authentication to processing, settlement and reconciliation.”
Jessica Cheney
The conversation repeatedly returned to the idea that the most effective fraud controls are often the ones that prevent a payment from being released in the first place. Verification of Payee, behavioral analytics, and stronger monitoring earlier in the customer journey are becoming increasingly important as financial institutions look to stop fraudulent payments before funds move.
Verification of Payee provides a good example. Giorgio Andreoli explained that the European scheme now processes more than one billion transactions every month, while maintaining an error rate below 1%. Just as importantly, fraudsters are trying to work around the control, a strong indication that it is having an impact.
“In the context of certain types of scams, e.g., an invoice fraud, receiving a NO MATCH signal during the Verification of Payee phase will send a warning to the victim and make them realize that the scam is ongoing. It is proof that this works.”
Giorgio Andreoli
For me, this is where the conversation becomes especially important. Recovering funds after a fraud event remains difficult, particularly in an instant payments environment. The greatest opportunity lies in identifying risk earlier, before a payment is released and before a customer becomes a victim.
Connecting Fragmented Risk Signals
A second theme centered on the way institutions organize fraud and risk management.
Most banks and payment providers have dedicated teams responsible for fraud, cyber security, AML, compliance and payments operations. Each performs an important role. The challenge is that fraudsters do not operate within those same organizational boundaries.
“Fraud has become end-to-end. If you think about identity verification, authentication, fraud detection, scam detection, mule detection, and post-transaction intelligence, you really need to connect the dots and the risk signals end-to-end.”
Genevieve Gimbert
That challenge becomes even more acute in a real-time payments environment. As payment speeds increase, the time available to identify, assess and respond to suspicious activity continues to shrink.
The industry does not have a shortage of data. Financial institutions already hold vast amounts of customer, transaction, and behavioral information. The problem is fragmentation: those data sets sit in different systems and teams, making it difficult to turn isolated signals into actionable intelligence. This is increasingly where institutions are focusing their attention. The conversation is shifting from “do we have enough data?” to “how do we connect the data we already have?” In many organizations, fraud, cyber, and compliance teams are all seeing indicators of the same threat, but through different systems, different processes, and different lenses.
Genevieve Gimbert emphasized that financial institutions need to combine signals from identity verification, onboarding, authentication, login activity, transaction monitoring and post-payment intelligence if they want a more complete view of risk. Looking at any one signal in isolation may only tell part of the story.
I agree with Genevieve as one of the biggest shifts I have seen in recent years is a growing recognition that adding more controls does not automatically reduce risk. In many cases, greater value comes from connecting existing controls, improving visibility and enabling teams to act on intelligence more quickly.
AI Raises the Stakes
No conversation about payment security in 2026 is likely to avoid the topic of AI, but what made this discussion interesting was how practical it became.
There was a broad agreement that AI is creating opportunities and challenges at the same time. Financial institutions are using AI to improve risk detection, analyze behavior patterns, and support decision-making. Meanwhile, fraudsters are using the same technology to scale attacks through false identities, deepfakes, voice cloning and increasingly sophisticated social engineering techniques.
“AI is industrializing fraud and scams.”
Genevieve Gimbert
Rather than focusing solely on the technology itself, the discussion centered on governance, accountability and oversight. Banks and PSPs need to think carefully about how AI is designed, deployed, monitored and controlled throughout its lifecycle.
What I hear most often from customers is that they are not looking for AI for AI’s sake. They are looking for practical ways to improve decision making, reduce false positives, and identify risk earlier. The organizations seeing the greatest value are typically those using AI to connect intelligence across different stages of the customer journey rather than treating it as another standalone capability.
AI will become an essential part of the security toolkit because it can uncover patterns, relationships and risks that would otherwise remain hidden within large volumes of data. But AI cannot operate alone. It must be combined with human expertise to interpret nuance, challenge decisions and maintain accountability.
Collaboration Beyond the Institution
Another strong theme from the session was that end-to-end security cannot stop at the boundaries of a single organization.
Fraud networks already operate across banks, payment providers, telecom operators and jurisdictions. Defending against them increasingly requires the same level of coordination.
Giorgio Andreoli outlined the European Payments Council’s work on FRIDA, a proposed fraud information-sharing framework designed to distribute fraud intelligence across participating payment service providers. Alongside Verification of Payee and emerging digital identity initiatives, it represents another layer of collective defense.
He positioned these initiatives as complementary rather than competing controls. Common rails and shared intelligence allow institutions to collaborate on security while continuing to differentiate in the services they provide.
Genevieve Gimbert welcomed the European Payments Council’s approach and pointed to the need for similar progress elsewhere. She encouraged industry participants, including policy makers, to support stronger data-sharing initiatives. While the US Financial Crimes Enforcement Network’s Section 314(b) framework provides a mechanism for financial institutions to share information, she argued that it does not go far enough. She also called on the Financial Action Task Force and Financial Stability Board to help strengthen and standardize the legal pathways for sharing fraud intelligence and payments data.
Fraud is a genuinely ecosystem-wide challenge, and that creates an opportunity for institutions to strengthen one another’s defenses. By sharing intelligence safely and efficiently, banks and their partners can gain a broader view of mule networks, threat vectors and emerging fraud patterns than any organization could develop alone. Shared intelligence is becoming an essential complement to the controls operating within individual institutions.
Cross-border collaboration can extend those benefits across the global payments system. Fraud prevention objectives must be balanced against privacy requirements, data protection obligations and differing regulatory frameworks - such as GDPR and data sharing across borders in Europe.
Even so, there is growing recognition across the industry that this challenge cannot be solved by institutions acting alone. As payments become more instant and global, the ability to share intelligence quickly and safely will become a competitive necessity rather than simply a regulatory discussion.
“Banks, payment providers, telecom companies and customers play a role in scams. It is really important to think about how you can prevent the loss from the beginning, versus allocating the blame.”
Genevieve Gimbert
From Isolated Controls to Connected Security
If there was one message that surfaced consistently throughout the discussion, it was that end-to-end security is no longer an aspiration. It has become a requirement.
Financial institutions have spent years strengthening individual controls, and those controls remain essential. Strong authentication, transaction monitoring, sanctions screening and fraud detection will continue to play a critical role. However, fraudsters move across channels, systems, and stages of the customer journey. Institutions need a similar ability to connect information across those same environments.
That starts with understanding the customer and payment journey in its entirety. It means identifying where controls exist, where visibility is limited and where risk signals become fragmented between systems or teams.
Genevieve Gimbert recommended beginning with a detailed mapping of the customer and payment journey. Financial institutions need to understand not only which controls are in place, but whether those controls appear at the right points in the process and whether they work effectively together.
Accountability also matters. Specialist teams will continue to own fraud, cyber security, compliance and AML responsibilities, but there is a growing case for clearer ownership of fraud outcomes across the broader payment journey.
“End-to-end transaction security cannot be delivered through a single technology, control or organization. We need clear institutional ownership and connected intelligence across the payment lifecycle globally.”
Jessica Cheney
Looking ahead, the most successful financial institutions will be those that stop treating fraud as a series of individual events and start viewing it as a connected journey. Most financial institutions already possess many of the controls, technologies and data sources they need. The greater challenge is bringing them together in a way that provides a more complete view of risk.
That was the strongest message I took from Sibos: in a world of real-time payments, AI-enabled fraud and growing cross-border complexity, security cannot remain confined to a single team, technology or stage of the transaction. Financial institutions must connect intelligence across the entire customer and payment journey, and they need to start now.
Continue the Conversation:
Share