Document Fraud Scaling Faster Than Manual Controls 

By Sean Quagliani, Fortiro CEO & Cofounder

This week’s media coverage of suspected fraudulent loan origination has put a spotlight on a reality many risk teams already live with. The most damaging frauds rarely arrive with obvious faults. They arrive as clean looking documents and images that move quickly through busy workflows, especially when volume is high and decisions require a quick turnaround. 

This is not a niche problem, and it is only getting bigger. In the past year alone, document fraud continues to evolve. Fraud template activity has risen sharply, accompanied by a steady increase in the use of AI tools for document fabrication. It is a strong signal fraudsters have moved from improvised forgery to industrialised production. Manipulated bank statements, payslips and supporting evidence are now built, packaged and repeated at scale. 

At Fortiro, we see the downstream cost of that shift every day and across our global customer base. Fortiro prevents, on average, A$57 million in fraud losses every month simply by detecting fraudulent documents and images before they become an exposure.  

The point is not that fraud is everywhere. The point is that fraud is now digital, fast and repeatable, and it targets the front end of high-value processes where a single false acceptance can have an outsized impact. 

The story behind most modern document fraud is simple. Digital documents can be altered in minutes. AI tools can generate believable income evidence. Invoices can be intercepted and have bank account details swapped without leaving obvious visual traces. PDF metadata is easily stripped. Composite images can be assembled from multiple sources. These techniques do not require deep expertise anymore. They require access to tools, templates and distribution. 

That is why the traditional model, a trained reviewer scanning for visual inconsistencies, has been outpaced. Manual review still matters, but without being guided and enabled by technology it is no longer a reliable sole control when forgeries are clean and volumes are high. In many portfolios, the most convincing fakes are the ones that have already passed the human eye. 

The broader fraud backdrop reinforces the scale of the challenge. The Association of Certified Fraud Examiners in the US analysed 1,921 investigated cases worldwide in its latest report and recorded more than US$3.1 billion in confirmed losses. That figure is not a theoretical estimate. It is tied to cases that were formally investigated, which means it is almost certainly a very conservative representation of the true impact. 

So, what needs to change? 

• First, organisations need to treat document and image verification as a first-class control, not an administrative step. If the evidence that supports approval is compromised, every downstream safeguard inherits that risk. 

• Second, “good” now means testing how a document is constructed, what it says and how it looks. That includes forensic signals that surface manipulation, inconsistencies across fields, artefacts of editing, and patterns that indicate templates and repeat use. It also means producing decision-ready outputs that improve auditability and speed, rather than forcing teams into more manual effort. 

• Third, controls must be consistent end-to-end. Fraud does not respect channel boundaries. It follows the path of least resistance across digital onboarding, covering assisted (brokers, introducers, bankers) and direct (online/digital) consumer and business channels. If one pathway is weaker, it becomes the preferred route. 

• Finally, a cultural shift is required. Fraud prevention is often framed as a cost. In reality, it is a trust and resilience function. It protects customers, partners, the business and in many regards, society. And the right control can save more money than it costs. 

The industry has reached an inflection point. Fraudsters have modernised. Now our controls have to modernise, too. 

Find out more about how Fortiro can help you manage document fraud. Join our webinar on Tuesday 16 March 2026, where we discuss the risk function of a high-growth lender in that context.