An AI document forensics solution improved fraud detection at an online mortgage broker by 30 percent, alongside reducing workload for investigators.
Online mortgage broker Habito announced a 30 percent improvement in mortgage fraud detection, following the implementation of an AI document forensics solution.
Alongside this improvement in detection, Habito also noticed a reduction in the amount of time financial crime investigators needed to solve escalated cases. On average, investigators spent 52 minutes less per case with the document forensics solution.
SEE ALSO: Machine Learning Isn’t An Instant Fix For Fraud
“Resistant AI identifies manipulated documentation far faster and far more accurately than we humans can and also brings us to conclusions faster and with more confidence,” said Matthew Willis, financial crime investigator at Habito. “This has broadened our horizons to the risks that exist with digital documentation, whereas before, we were stumbling in the dark. They have helped us to drastically reduce both the time it takes to catch fraud and the amount of fraud that makes it past us to lenders.”
Resistant AI’s solution was integrated into Habito’s mortgage application process, with workflow triggers tied to several parts of the process. The machine learning algorithm is able to spot document forgery, serial fraud, synthetic identities, bots, account takeovers, and money laundering.
In cases of clear fraudulent manipulation, the application was automatically rejected. The AI is also able to tag documents with warning or high risk triggers, which are then viewed by specialist investigators. The automatic rejection of clearly forged documents reduced the amount of documents ending up in investigator’s mailboxes.
“Habito has established a strong reputation for innovation and frictionless customer experience, and the implementation of Resistant AI Document Forensics underlines their commitment to the highest standards of performance and efficiency,” said Martin Rehak, founder and CEO at Resistant AI.