62% say Gen AI is behind invoice fraud

Gen AI is increasing difficulty in distinguishing between legitimate and fraudulent documents, as attacks grow in both volume and sophistication. This was uncovered in Basware's 'The Rise in AP Fraud' report, surveying 100 global business leaders and CFos on the exposure companies have to financial fraud, as well as the protection measures they have in place.

In addition to this, 90% of organisations lack dedicated fraud prevention teams which forces accounting staff to juggle fraud management beside their regular responsibilities, often straining resources and increasing vulnerability. 

Fraud is becoming increasingly frequent, sophisticated, and damaging for businesses. In 2024, seven out of 10 organisations (68%) reported fraud attempts, with 62% noting these attempts had worsened over the past year. Last year aloen, fraud schemes resulted in global losses of $485 billion.

Remote work and decentralised operations are seen as key contributors of this increase, as identified by a quarter (25%) of respondents. By weakening internal controls, remote work exposes accounts payable (AP) departments to greater vulnerability, making it harder to prevent fraudulent activities. 

At the same time, approximately a quarter of respondents also point to rising transaction volumes (29%) and the expansion of the supplier base (25%) as critical risk factors. With more invoices to process and a larger number of suppliers to manage, AP teams face greater complexity, which increases the likelihood of errors or fraud.  

Gen AI being used to create illegitimate invoices

AP departments are on the frontlines of fraud detection, while also managing supplier payments, leaving them overwhelmed by the complexity and growing volume of fraudulent attempts. 

Increasingly sophisticated tactics, such as fake invoices, deepfake media, and phishing scams - driven by gen AI - have made attacks more convincing and harder to detect. Consequently, two-thirds (62%) of businesses now cite generative AI as a key driver behind the surge in AP fraud.  

Tools such as ChatGPT enable the mass creation of communications and fake invoices, increasing risks and delays in resolving legitimate disputes. When handled by multiple disconnected systems, information becomes fragmented, making patterns harder to spot and allowing fraud to slip through undetected. 

A high-impact case of AP fraud involved a man from Lithuania, who defrauded Facebook and Google out of more than $120 million by posing as a legitimate supplier. Using fake invoices, he secured $99 million from Facebook and $23 million from Google. The fraud underscored how vendor impersonation and fake invoices can cause massive financial losses without robust AP safeguards, even for large enterprises.  

According to the survey, 90% of organisations lack dedicated fraud prevention teams, causing understaffed and under-resourced AP teams to prioritise speed over detailed checks, creating vulnerabilities. Meanwhile, reliance on paper-based processes and inadequate statement reconciliation leaves teams exposed to errors, inefficiencies, and fraudulent activities. 

"Manual processes are inherently slow and prone to errors, making it difficult to match invoices, track approvals, or identify duplicate payments – creating exploitable gaps for fraudsters, who are now using GenAI. For overburdened AP teams, these create the perfect storm of challenges, leading to operational bottlenecks, strained supplier relationships, and lost cash flow," said Tom Santacroce, Global VP of AP Assurance."Forward-thinking organisations are reimagining fraud, overpayment and risk prevention through AI and automation that protects against increasingly complex financial threats. Remote work has weakened traditional security, requiring secure systems and innovative solutions like decentralised finance and blockchain for transparent vendor transactions. With increasingly sophisticated fraud tactics on the uptick, organizations must prepare for stricter AI and compliance rules. Proactive fraud prevention today not only mitigates risk, but also future-proofs operations against costly penalties." 

Combatting fraud

AP fraud is one of the most targeted types of fraud for businesses. The survey found 28% of organisations cite a lack of AP automation to help tackle fraud. Reliance on manual processes and outdated tools leaves businesses ill-equipped to manage complex, high-volume transactions, creating critical gaps for fraudsters to exploit.  

One example of how automation has strengthened AP fraud prevention is in KION, who were previously vulnerable to fraud risks due to manual invoice processing. To address these challenges, KION partnered wikth Basware to shift to automated AP processes driven by AI, resulting in reducing manual errors and minimising fraud opportunities. With over 90% of its spend now controlled, KIOn has streamlined its operations and reduced the risk of errors or AP fraud.