Fraud is no longer an edge case for finance teams, it’s a daily operational risk. According to Trustpair’s 2026 U.S. Fraud Report, 71% of companies reported an increase in AI-powered fraud attempts over the past year, yet nearly half still rely on manual validation methods to protect payments. At the same time, regulatory pressure is rising, payment rails are accelerating, and the operational and human costs of fraud continue to grow.
To bring these findings to life, we spoke with Lee-Ann Perkins, Head of Global Treasury at Ankura Consulting, about how fraud is evolving in practice, and what treasury teams must change to stay ahead. Download the full report for more details.
71% of companies have seen an increase in AI-powered fraud. From a treasurer’s perspective, how has AI changed fraud in practice?
From a treasury perspective, fraud is no longer an exception, it’s a daily operational risk we manage constantly. That hasn’t changed. What has changed is the scale, the speed, and especially the credibility of fraud attempts.
What used to be occasional, low-quality scams are now frequent, well-timed, and extremely convincing. AI has removed many of the obvious red flags we used to rely on, like poor spelling or awkward formatting. These attempts look legitimate, they arrive at the right moment, and they’re designed to deceive experienced professionals.
The biggest concern for me is the imbalance we’re seeing. Fraudsters can innovate every day, it’s their full-time job. Meanwhile, many companies are still relying on manual, reactive controls. Fraud today is no longer opportunistic; it’s industrialized, automated, and fundamentally different from what most of us grew up with in our treasury careers.
Despite rising AI-powered fraud, 48% of companies still rely on manual validation methods. Why do these processes persist?
Manual controls persist because they’re familiar, inexpensive, and historically they worked. In a slower environment, callbacks and email confirmations were often enough to catch basic scams.But those methods don’t scale against AI-driven fraud or real-time payment rails. Today, treasury teams are trying to apply yesterday’s tools to today’s threats.
Modernization is not trivial. It takes time, budget, cross-functional coordination, and buy-in from IT, AP, accounting, and leadership. Treasury teams are small, and system change is complex. But the reality is clear: if we want a chance of success, we need to meet technology-driven fraud with technology-driven defenses.
In your experience, when do manual controls fail most often?
They fail under speed and volume. End-of-month pressure, urgent payments, year-end activity, and high vendor turnover, these are exactly the conditions fraudsters exploit.
Humans are most vulnerable when we’re rushed or overloaded. AI-powered fraud is designed to take advantage of that. Treasury teams are hired for speed and precision, but that same urgency can be turned against us.I often say we need to prioritize security over immediacy. A deliberate, verified payment is always better than a rushed one that exposes the company to fraud.
Nearly half of companies aren’t aware of Nacha’s 2026 requirements, and only 57% feel confident they’ll meet the deadline. Why is readiness so difficult?
That statistic is concerning, but I don’t think it reflects a lack of understanding of the rules. It’s more about execution.
Compliance depends on many moving parts: systems, data quality, vendor cooperation, and coordination across teams. When controls are fragmented or data isn’t continuously validated, meeting regulatory requirements becomes very difficult in practice.
Regulations often move faster than operational change, especially in large organizations. Treasury teams need top-down support, budget, time, and resources. Fraud prevention and regulatory compliance shouldn’t be a battle internally; they should be treated with the same urgency as audit and accounting controls.
How do regulations like Nacha and SOX change the way treasury teams think about fraud prevention?
They force a shift from detecting errors after the fact to preventing them upfront.
SOX pushed us to embed controls directly into payment processes. Nacha reinforces that expectation by requiring defensible, documented proof that companies are validating payee information before payments are sent.
Good-faith effort is no longer enough. Companies must demonstrate consistent monitoring, training, documentation, and prevention. That requires automation and auditable processes, not just policies on paper.
Our study shows a major confidence gap: 77% of companies trust their vendor data, but only 32% validate it continuously. Why?
It almost feels like companies think confidence itself is a control, but it’s not. Fraud thrives in the gap between what we believe is correct and what we actually verify.
On paper, controls often look strong. But fraud tests controls in real time. Payments go through successfully until they don’t, which creates a false sense of security.
Without continuous validation, teams are trusting static data in a very dynamic risk environment. And because many stakeholders touch data across the P2P process, this becomes a shared responsibility, not something treasury can manage in isolation.
With 75% of companies using or planning to use instant payments, how does speed change fraud prevention?
Instant payments fundamentally change the role of treasury. We’re moving from detect and recover to prevent and block.
With real-time payments, reaction windows shrink from hours to seconds. Detection after the fact is no longer viable. Fraud prevention has to happen before the payment is released.
Treasury teams are becoming gatekeepers of enterprise risk. Controls must operate in real time, not in batch reviews, and that again comes back to automation and clean data.
Fraud also has major operational and human consequences. How does a fraud incident impact teams?
Fraud is not just a financial loss, it’s an organizational crisis. A million-dollar fraud doesn’t cost just a million dollars; it costs months of disruption.
Teams are pulled into investigations, operations slow down, trust is strained with vendors and customers, and sometimes careers are impacted. There’s also burnout: people spending days fixing something that should never have happened.
If companies don’t equip employees with the right tools, we’re setting them up to fail. Modern fraud is not a people problem, it’s a systems problem.
Looking ahead, what should treasury teams change first to stay ahead of fraud?
Modernization is non-negotiable. Fraud will continue to increase: the question is whether companies evolve fast enough to withstand it.
Treasury teams need to move away from reactive, manual controls toward preventive, automated ones. Real-time decisioning, continuous monitoring, and pattern recognition must replace daily reviews.
When systems are in place, the biggest benefit is removing opportunities for error before money moves. Systems can be fixed, and that’s where companies should focus their investment, time, and attention.
Final thoughts
Fraud today is fundamentally different from what we’ve seen in the past. AI has changed the scale of the problem, and companies must respond with comparable tools. Awareness is no longer enough: control, automation, and prevention are now essential.