Industry Thoughts

How to identify fake loan documents automatically

April 24, 2021

minute read

  • Ryan Gross
    Head of Sales, Inscribe

Using document fraud review software leads to fewer loan write-offs

Today’s online lenders ask applicants to upload bank statements, pay stubs, tax forms, and other financial documents to assess creditworthiness. Our data shows that 5% of all financial application documents submitted in an online channel have been manipulated. Further, our research suggests that a fraudulent application document leads to a loan write off rate of 60% on average. Document manipulation is highly correlated with loan write-offs.

How do you identify fake loan documents?

The old way of detecting document fraud

The old way of identifying document manipulation is using the human eye. The human eye can spot obvious inconsistencies in text style, spacing, alignment, and color. Further, a well trained, experienced fraud analyst will review the metadata properties of a document looking for suspicious traces of manipulation such as PDF editing software like Photoshop or an untrusted document author.

The old way of detecting document fraud is limited to the human eye and metadata property analysis. Well trained analysts can conduct a thorough investigation of an individual document in 5-10 minutes.

Problems with the old way of identifying fraudulent documents

A quick Google search will result in hundreds of websites where a fraudster can buy a fake financial document. The human eye cannot reliably identify manipulation in these documents. So how do you combat the rampant use of entirely fake, online purchased financial documents?

The new way of detecting document fraud

Ask your best fraud analyst what properties should be contained in the latest version of a Bank of America personal checking bank statement. They should be able to tell you the font style, font size, latest logo placement, the expected URL, the expected document author, and the alignment of name/address within the top left corner of the document.

It’s unrealistic for a fraud analyst to memorize all the metadata properties expected for an individual bank statement yet alone the statement properties for the thousands of different banks your applicants use.

We’re just scratching the surface on the advantages software has in analyzing the legitimacy of documents when compared to the human eye. With Inscribe, you’ll receive an analysis in 10 seconds, of the metadata, pixel level information, and file history of a document. Inscribe’s software is night vision goggles for your fraud team.

The cat and mouse of document fraud

Document fraud is the oldest fraud technique in a fraudster playbook and it’s not going away any time soon. As newer and more sophisticated forms of document manipulation emerge, fraud teams need technologies who specialize in racing fraudsters to the next method of fraud detection.

As a centralized technology, Inscribe empowers your team with learnings we gain across our customer base. As we see new fraud methods pop up, we’re able to quickly innovate so your team can stay ahead of the latest fraud trends.

Easily detect document fraud with Inscribe

Lenders can fight back against document fraud with the right technology. Inscribe uses rules-based fraud detection capabilities and machine learning to generate a full analysis of the legitimacy of an application document.

Inscribe’s automated fraud analysis enables your team to make quicker and more precise decisions. The Inscribe product provides more efficient and ultimately automated document fraud reviews leading to fewer loan write-offs. Inscribe filters out the bad so your team can focus on the good.

Contact our sales team today to learn more.

Start your AI journey today

See how Inscribe can help you unlock superhuman performance with AI Risk Agents and Risk Models.