Verity Credit Union

Verity Credit Union, a CDFI with substantial assets, relied on generic scoring models that left qualified members — particularly those from protected classes and middle-lower credit tiers — without access to credit. Biased and inaccurate decisioning methods led to inconsistent outcomes that conflicted with Verity's mission of fair and sustainable lending for underserved communities.

Verity implemented Zest AI's AI-automated underwriting platform, integrating it with their existing LOS. The solution applied machine learning models trained for fairness to auto loan, personal loan, and credit card decisioning, enabling the credit union to lend deeper across the credit spectrum with confidence and gain lending intelligence reporting for ongoing performance monitoring.

Approval rates for protected classes rose significantly: substantially for individuals aged 62+, meaningfully for African Americans, 375% for Asian Pacific Islanders, notably for women, and considerably for Latinos. Automated instant approvals increased 100% for auto loans, substantially for personal loans, and 84% for credit cards. The improved lending confidence also qualified Verity for additional CDFI-funded opportunities, enabling new products like Sharia-compliant loans and a microgrant program.

• AI underwriting can simultaneously improve approval rates for underserved borrowers and increase automation speed, resolving the false tradeoff between fairness and efficiency.
• CDFI lenders can use AI to expand mission-aligned lending and unlock additional grant eligibility by demonstrating responsible, data-driven credit extension.
• Seamless LOS integration is a critical adoption factor; Zest AI's native integration reduced friction and enabled faster deployment.