First American Data & Analytics, a division of First American Financial, has launched a new tool that allows lenders and investors to more precisely identify the risk of fraud or early payment default on both new mortgage applications or a portfolio.
The company’s Press release described AppIntelligence Score (AI Score) as a fraud pattern-recognition score that uses proprietary predictive modeling to pinpoint where fraud and early payment default risk are likely to occur. First American said that this targeting allows lenders to focus their reviews on the most at-risk loans while streamlining loan approvals and shaving operational costs.
Additionally, the fraud score technology features a “retro-scoring capability” that enables lenders to run a portfolio to identify which loans would have been flagged by the model. AI Score then produces a scale of scores from low to high risk, which is fed back the data into the model for continuous machine learning.
“The ability to analyze millions of alerts that underwriters have cleared, and then feed that information back into the model, is what makes the AI Score one of the most sophisticated models on the market,” said Robert Karraa, president of First American Data & Analytics. “By targeting the most at-risk loans, AI Score will help lenders streamline the number of review cycles and still uncover the highest risk areas of fraud.”
AI Score can also simultaneously run proprietary sub-models for risk, including synthetic identity, income, employment, early payment default (EPD), undisclosed debt and loan participant risk review.