Late-stage delinquencies are rising across all credit tiers, the average VantageScore dropped to 701, and more consumers are shifting into subprime. At the same time, leading institutions are rethinking how they make credit decisions, and machine learning is changing what's possible.
In this episode, Jay Budzik, SVP and Director of AI and Machine Learning Operations at Fifth Third Bank, explains how the same technologies that power voice transcription and lane-assist in your car are now being deployed to make more accurate and fair credit decisions.
In the full episode, Jay covers:
➡️ How upgrading to more accurate credit scores can simultaneously improve bank profitability and expand access to credit for overlooked borrowers
➡️ Why transparency and explainability have historically been the biggest challenge for machine learning in credit, and how the field has evolved to address it
➡️ What Fifth Third watches beyond credit scores, including consumer deposit data as a first-stop indicator of financial health
Plus, VantageScore's Susan Fahy and Atif Mirza break down the latest CreditGauge™ data for July 2025, including the average VantageScore declining to 701 as more consumers shift into subprime, mortgage, and auto loan delinquencies, posting the largest year-over-year increases, and a divergence emerging between secured lending softening and unsecured credit growing.
Download the full CreditGauge™ analysis at vantagescore.com/lenders/credit-gauge.
Updated monthly, VantageScore’s CreditGauge LIVE features the latest newsmakers and insights into consumer credit health.
Download the full CreditGauge analysis for free at vantagescore.com/lenders/credit-gauge.