Case Study: Millbrook Business Finance
How Millbrook adopted machine learning automation to protect its finance network from high-risk applications.
We’re trusted by the industry-leading companies
per month saved in application reviews whilst increasing effectiveness and accuracy.
Over 300 risk tests on documents prevent finance from being issued to fraudsters.
of suspicious documents submitted confirmed high-risk.
Since implementing Fraud Finder, Millbrook Business Finance has significantly boosted our ability to detect fraudulent activity across our network. This enhanced capacity has played a vital role in increasing safety for everyone involved, ensuring our lending partners are better protected from high-risk applications and saving them valuable time and resources.
Fighting fraud for Millbrook with AI document analysis
Fraudfinder enables Millbrook to authenticate financial documents, safeguarding against Fraud efficiently
Millbrook Business Finance is a UK-based award-winning financial services provider dedicated to helping small and medium-sized businesses access tailored funding solutions. Millbrook provides flexible commercial financing options, helping reduce common funding barriers. By offering personalised solutions crafted to meet specific requirements and business types, they make business finance more accessible to everyone.
Millbrook is on a mission to simplify the business finance landscape, offering a quick and straightforward application process to a wide range of sectors with funding options ranging from £10,000 to £25 million. Millbrook champions the values of transparency, commitment and dynamism.
Fraud is the most commonly experienced crime in the UK. It accounts for over 40% of all crime in England and Wales and costs an estimated 6.05% of the global GDP. CIFAS reports that 214,882 cases were filed to the NFD (National Fraud Database), between January and June 2024. A 15% increase on the year before. On average, businesses lose approximately 5% of their annual revenue to fraud. In the UK, the average cost of dealing with fraud under £1,000 can amount to £19,793 in losses.
Despite Millbrook's excellent ability to assess documents for fraud using human intelligence alone, fraudsters are continuously evolving and developing new tactics. The emergence of AI technologies has transformed the landscape by enabling the creation of fraudulent documents that closely resemble authentic ones. Machine Learning automation is no longer a nice-to-have, but an essential.
For the last year or so, Millbrook Business Finance has been employing Fraudfinder, our sophisticated document authentication tool, to detect and prevent fraud in financial documents via a secure web portal. Our system instantly validates document authenticity, flags modifications, and automates the review process, reducing assessment time while minimising human error.
Utilising advanced techniques like image analysis, metadata extraction, and QR/barcode scanning, Fraudfinder streamlines Millbrook's document analysis processes, enabling quicker approvals while reducing risk. In cases where additional clarification is required - Millbrook can rely on Fraudfinder’s skilled Operations team, who use their expertise to provide further recommendations and support when necessary.
Below, we show examples of three cases that Fraudfinder correctly identified as fraudulent. The total loan value of these three cases, which would not have been identified without Fraud Finder, is £20,000.
Example 1: Fraud in a Santander bank statement
Fraudfinder works by evaluating each document against an extensive, adaptive database of authentic institutional document templates. Our first fraud example was altered to such an extent that it no longer aligns with the expectations of a Santander statement. The presence of suspicious fonts, highlighted in green and blue, indicates modifications to the document’s dates and end balance. Additionally, the font used for the applicant’s name, sort code, and IBAN is also flagged as atypical for Santander. This pinpoints exactly on the document where tampering has taken place.
The second example showcases a fraudulent Monzo bank statement. The file was edited just thirty-five minutes after being downloaded, leaving behind multiple document versions that expose a trail of deleted and overwritten data. This advanced feature functions as an "x-ray," revealing edits entirely undetectable to the human eye.
In this case, the original data, highlighted in blue, was replaced with the modified text shown in orange, where the applicant altered the original balance from £3.59 to £12,933.46. Changes were made to the total deposits and two significant incoming transactions. It’s a HUGE jump to claim you have almost £13,000 in your account when you really have less than a fiver!
In our third example of a Barclays business bank statement, Fraudfinder has identified multiple instances of document tampering. Notably, this document was produced using a commonly available PDF software that allows users to create, edit and convert documents. The document was edited a full ten months after it was generated by the bank! The font highlighted in green indicates that the statement's year has been altered. Given that the document was initially created in April 2023, but subsequently edited in February 2024.
The applicant attempted to present an outdated bank statement as a recent one. The risk associated with an outdated bank statement is huge, you may as well not show one at all.
This loan was valued at £3,500.
By integrating Fraudfinder into its operations, Millbrook Business Finance has substantially enhanced its fraud detection capabilities, helping to safeguard both the organisation and its clients from the evolving tactics of fraudsters. The system’s intelligent automation not only speeds up the approvals process but also reduces human error and provides valuable operational insights. In turn, Millbrook’s reputation for transparency, commitment, and dynamicity remains robust, and its clients benefit from a safer, more reliable pathway to securing the funding they need.
Any finance company can trial Fraudfinder for free and start protecting their loans with automation just as Millbrook Business Finance have. Either click here to book a demo, or email harry.foster@fraudfinderai.com.
About
Fraud and risk in business loans
Features applicable:
Business bank statement validation
Previous versions fraud test
Barcode decoding
Font analysis
Structure, layering and metadata tests
See Fraud Finder in action
Instant fraud detection with the click of a button...
Looking to automate your fraud risk process?
Book a demo today...
We'll provide a short no-obligation demo of each element of our fraud tech, book below:
