As payment fraud grows more sophisticated, organizations are struggling to keep pace. Trustpair’s 2026 Fraud Report shows that 71% of U.S. companies saw an increase in AI-powered fraud last year, while only 32% continuously validate vendor bank data. To explore what these findings mean from a banking and data-network perspective, we spoke with Gloria Wan, Executive Director at Kinexys by J.PMorgan, about how AI, data sharing, and regulation are reshaping fraud prevention – and where companies are still falling behind.
AI-powered fraud is rising sharply. From your perspective, how is AI changing the scale and sophistication of payment fraud?
What we’re seeing is a significant increase in the sophistication and targeting of fraud attacks. Compared to even two years ago, fraudsters now have access to much richer data sets and far more advanced tools. That allows them to coordinate attacks using multiple data points that all appear consistent and legitimate.
This makes fraud much harder to detect. Traditional indicators that used to work well are becoming less effective, and in areas like synthetic identity fraud, it’s increasingly difficult to determine what is real and what is fake. The result is fraud that looks “close enough” to legitimate activity to pass many existing checks.
The report shows that 58% of organizations believe fraudsters are evolving faster than humans can respond. What role do shared networks and real-time data play in closing that gap?
Shared networks and real-time data exchange are absolutely foundational, but the industry still has a long way to go. To be effective, fraud prevention requires access to highly sensitive transactional data – often across borders – and that raises real challenges around privacy, governance, and consistency.
What’s critical is building secure, permissioned data connectivity that extends visibility in a controlled way. Before we even talk about AI, we need a strong data foundation: consistent data, global coverage, and the ability to operationalize that information.
We also need to move beyond isolated data points. Today, fraud detection depends on proving relationships between multiple data elements – building a complete data profile for each beneficiary and transaction. That collective intelligence is where the industry needs to head next.
Regulation is rising, but readiness remains uneven. Where do you see the biggest gaps with Nacha 2026 and similar initiatives?
We tend to see organizations move through different stages. The first is awareness, simply knowing that new regulations or tools exist. Many companies are still stuck here with Nacha’s upcoming new rules.
The second stage is technical readiness: having the ability to connect to APIs or adopt new tools. The third, and most overlooked stage, is operational effectiveness. Even when companies receive new data, many don’t know how to act on it or embed it into decision-making processes.
We’ve seen corporates say they’d rather opt out of receiving certain validation data – for VoP for example – because they’re not ready to handle the operational consequences. That’s a major challenge. Regulation only works if organizations can actually use the information it requires.
Our data shows high confidence in vendor data quality – but weak validation practices. Why this disconnect?
This gap comes down to education and visibility. If organizations don’t know what “good” looks like, they assume their current processes are sufficient – especially if they haven’t yet experienced a major fraud incident.
Fraud prevention often becomes reactive. Companies realize there’s a gap only after something goes wrong. Another major factor is internal silos. Procurement, payments, and treasury teams may each believe their own controls are strong, but the end-to-end process isn’t connected.
Those gaps between teams are exactly where fraudsters look to exploit weaknesses.
Faster payments are becoming the norm. How does that change fraud risk?
Real-time payments significantly raise the stakes. As settlement speeds increase – especially in B2B use cases – the window to detect and stop fraud shrinks dramatically.
Looking ahead, this challenge will only grow. New payment rails, including blockchain-based payments, can reduce transparency even further and make recovery almost impossible. That’s why pre-payment validation becomes so critical.
Regardless of the rail – ACH, wire, real-time, or blockchain – you still need to know who you’re paying. Beneficiary validation and data verification remain the foundation of fraud prevention.
The report shows companies still prioritize training over automation. What’s your view on that balance?
Training and automation are both essential, but training alone is not enough. Many organizations are hesitant about automation because they worry about cost, implementation time, and uncertainty around ROI.
There’s also a lack of clear market data proving the effectiveness of some technologies, which makes companies cautious. But fraud risk isn’t waiting for perfect evidence. Organizations need to start the journey – even incrementally.
Training builds awareness, but automation is what enables consistent, scalable protection against increasingly complex fraud tactics.
Any final thoughts as organizations prepare for 2026?
Education remains the biggest gap. The industry needs clearer guidance on best practices – not just what regulations require, but how to implement controls effectively.
Fraud prevention is no longer a single-team problem. It requires collaboration across banks, corporates, regulators, and technology providers. The organizations that invest in data foundations and connectivity now will be far better positioned for what’s coming next.
Request a demo to learn more about Trustpair, the ultimate fraud prevention solution.