Rising Risk: Confronting Modern AP Fraud Threats

Fraud continues to pose a major challenge to accounts payable (AP) teams. Research shows that 68% of organizations encountered at least one fraud attempt in 2024, with AP fraud taking a variety of forms, including phishing attacks, account takeovers and invoice fraud. Moreover, fraudsters are rapidly growing more sophisticated as they gain access to fraud weapons fashioned with the help of artificial intelligence (AI). In recent years, AI-generated deepfakes and impersonations have become top threats, with nearly two-thirds of companies saying that generative AI (GenAI) is making fraud more dangerous than ever before.

Against these attacks, the manual tools and defenses of traditional AP, though still wielded by most firms, are proving woefully inadequate. When it comes to fraud, however, AI is a double-edged sword. As bad actors increasingly leverage the technology for fraudulent aims, AP departments are fighting back with AI-driven tools of their own. AI-powered fraud detection and automated systems reduce human error, enhance security, ensure regulatory compliance and provide better visibility and control. These systems are making it more possible than ever to meet — and defeat — fraud wherever and in whatever form it arises.

• A Growing Menace: AI-Generated Fraud
• The Inadequacies of Manual Fraud Detection
• Fighting Fire With Fire: AI-Driven Anti-Fraud Strategies
• Shifting From Manual to Automatic in Fraud Prevention

A Growing Menace: AI-Generated Fraud

Fraud is a constant battle for AP departments, but GenAI has introduced techniques like deepfakes and impersonations, making fraud even more dangerous.

GenAI has revolutionized fraud techniques — and risk.

A Basware study confirmed that 62% of CFOs see GenAI as a primary contributor to the rise in AP fraud. Another 34% said that keeping up with these evolving techniques is the top challenge AP departments face, especially for firms relying on manual fraud prevention processes.

As a direct result, fraud’s financial impact reached unprecedented levels last year.

The Trustpair research found that of the U.S. companies targeted by fraud in 2024, 86% lost money. Nearly half (47%) lost more than $10 million. This represents a 136% increase in financial loss due to fraud for U.S. companies between 2023 and 2024. This spike owes itself to AI as well, as automation of fraud vectors makes it possible for fraudsters to move more money faster than ever before.

While 90% of U.S. executives said they had confidence in their AP teams’ abilities to detect fraud attacks, the rising number of successful attempts indicates this confidence may not be warranted. One important reason for the failure of fraud detection, for example, is the continued heavy reliance on manual methods for this process.

The Inadequacies of Manual Fraud Detection

Despite the advanced cybersecurity technologies now available, most AP departments still rely on manual anti-fraud procedures, making for ineffective fraud detection and prevention.

Most AP departments still rely on manual methods for fraud detection.

Experts note that while manual fraud reviews and audits are useful for uncovering suspicious findings, they leave significant gaps during which fraudulent activity can go unnoticed, especially as transaction volumes surge and fraud tactics grow more advanced.

Human error is the most common source of payment fraud.

Manual processes, in fact, turn out to be the number-one factor in fraud, and the bigger the business, the worse it gets. Trustmi reports that more than one-fifth of companies are especially susceptible to payment fraud simply due to their high monthly invoice-processing volume. This is because a high volume of manual processing contributes heavily to human error — the largest single source of payment fraud, at 50%.

The sheer scale and complexity of financial transactions within large organizations make it nearly impossible for human reviewers to detect fraud in a timely manner or avoid slip-ups and oversights. Complicated tasks like invoice matching, approval tracking and identifying duplicate payments become increasingly difficult to manage accurately. Without real-time monitoring in place, unusual spending patterns or fraudulent activity can continue unchecked for weeks or even months, potentially leading to severe financial losses.

Fighting Fire With Fire: AI-Driven Anti-Fraud Strategies

As fraudsters embrace advanced technologies, AP departments must also deploy new methods. Automation and AI are key weapons for fighting new forms of cybercrime.

AP automation can significantly improve security.

Full automation of payment processes remains limited, however, with only 5% of organizations having completely automated their AP systems. Sixty-nine percent have adopted partial automation, though, a key step in fraud prevention. This growing trend reflects a shift toward digital transformation in financial operations, driven by the need for greater efficiency, accuracy and security in managing payments.

AI agents have proven effective at detecting fraud and human error.

AI-powered fraud detection systems can spot invoice anomalies, duplicate invoices and data entry errors that humans frequently miss. One example of AI anti-fraud integration is Routable’s incorporation of an AI agent into its AP automation platform. Initially, the platform’s optical character recognition (OCR) system scans invoices to create bills that are either automatically matched to a purchase order or predictively coded based on machine learning (ML) and historical data. The AI agent then scans for anomalies and errors and, if it finds anything questionable, alerts AP teams — thus stopping fraud before it happens.

“As fraudsters get more creative, it’s become impossible for humans to keep up,” Routable CEO and Co-Founder Omri Mor said in a press release. “[This solution] uses the power of AI to detect and prevent both invoice fraud and human error — giving AP teams superhuman visibility to protect the business.”

Shifting From Manual to Automatic in Fraud Prevention

Manual fraud prevention methods are becoming increasingly insufficient for AP departments due to their reliance on human oversight and susceptibility to error. Manual screening often fails to catch sophisticated fraud schemes such as those produced by GenAI because it is difficult for individuals to spot subtle anomalies across large volumes of transactions. Additionally, manual processes lack real-time monitoring and struggle to enforce consistent controls, both of which create opportunities for fraudsters to exploit weaknesses in the AP system.

Automation and AI-driven solutions offer a far more effective approach to fraud detection and prevention in accounts payable. Automated systems centralize all AP activities, enforce segregation of duties, and provide real-time monitoring, making it much harder for unauthorized transactions to slip through. AI can analyze vast amounts of transaction data, identify patterns and flag anomalies that would be difficult for humans to detect manually. These technologies also create comprehensive audit trails, streamline approval workflows and reduce the risk of human error. By leveraging automation and AI, AP departments can gain enhanced visibility, faster detection and stronger controls, making their accounting processes more secure against fraud.

The pace and sophistication of today’s fraud threats — especially those powered by AI — should be a wake-up call for every finance team. Relying on manual processes is no longer just inefficient — it’s dangerous. The time to modernize your AP operations is now, before your organization becomes the next cautionary tale.”