GUIDES & REPORTS
Closing the AP fraud gap
An SSON Essentials Guide, Sponsored by Medius
Most would agree, AP fraud is becoming more common. It's also getting smarter.
Criminals now use AI to generate convincing fake invoices, clone executive voices for payment authorization calls and run coordinated attacks that traditional AP controls simply aren't built to catch. 76% of organizations reported attempted or actual fraud in 2025, and most of them found out only after the money was already gone.
This SSON Essentials Guide, sponsored by Medius, gives AP and finance leaders the research, real-world context and practical framework to understand what they're up against, and what it takes to build controls that catch fraud before it becomes a loss.
What's inside:
- The four most common AP fraud risks with data on how widespread each one has become
- Why legacy AP environments leave organizations exposed, including the automation gap that's widening the window for fraud
- Real whistleblower accounts, exploring why employees often stay silent even when they suspect something is wrong, and what it costs when they do
- An AI capability vs. limitation matrix for fraud detection, where technology wins and where human judgment is still essential
- The three pillars of a fraud-resilient AP function: people, process and technology — and specific actions AP leaders can take across each
The threat has changed. Most AP controls haven't.
Seventy-five percent of organizations were hit by Business Email Compromise in 2025. Forty-four percent have been targeted by invoice fraud. And AI-generated deepfakes — voice and video impersonation attacks designed to solicit or authorize payments — have gone from an edge case to a mainstream threat.
53% of finance professionals have been targeted by deepfakes. 43% have fallen victim.
The controls most AP teams rely on weren't designed for this. Only 17% of organizations have automated more than half their finance processes. Nearly half have automated less than a quarter. That gap between the sophistication of today's fraud and the maturity of current AP controls is exactly what this report is built to help you close. See how AI-powered fraud detection can help you get ahead of it.
Why fraud goes undetected, even when people see it
Technology gaps are only part of the problem. According to Medius' Financial Census Survey, 1 in 4 finance professionals have no idea how much invoice fraud has cost their business. For those who do know, the average is $133,000 per year for US businesses and £104,000 for UK businesses. Those are only the losses someone actually spotted.
The Medius Accounts Deceivable podcast explored why the real number is probably higher. Through whistleblower accounts including Sherron Watkins, who uncovered the Enron scandal, the pattern is consistent: 56% of finance practitioners suspect internal fraud in their workplace. 81% say nothing. And 48% don't believe the legal system adequately protects those who do speak up. For AP managers and finance teams, that silence is as much a control risk as any technology gap.
Why you can't afford to wait
Only 17% of organizations currently use AI to detect payment fraud. But AI is already being used against them.
Closing the gap requires more than better software. It requires an integrated approach where intelligent AP automation and the right human controls work together, flagging anomalies before payments go out, monitoring vendor master data for unauthorized changes and creating the audit trails you need to respond fast when something goes wrong. Learn more about how Medius uses AI to protect AP.
This report shows you what a truly fraud-resilient AP function looks like, and what it takes to build one.
