Founder & CEO at FairPlay | 10+ Years of Applying AI to Financial Services | Architect of $3B+ in Financing Facilities for the World's Underserved
Sponsor banks operating received 35% of all regulatory enforcement actions in Q1, despite representing only 3% of banks. Thanks Konrad Alt at Klaros Group for this great visualization. A lot of that enforcement activity has been related to fair lending compliance. In many of these cases the main issue has been that sponsor banks were not able to demonstrate sufficient management and oversight of the fair lending risks arising from their fintech partnerships. Until these enforcement actions started coming down, common practice was for sponsor banks to rely on annual attestations by their fintech partners of compliance with non-discrimination laws. We’ve learned through recent consent orders though that the regulatory goal posts are now more clearly defined. Sponsor banks need to independently demonstrate: ⚫ Fairness: The credit policies, scoring models, and data used by their lending partners are fair. ⚫ Ongoing Monitoring: Disparities in underwriting, pricing, loan amounts, fraud detection, and other key decisions impacting consumers are actively tracked and addressed. ⚫ Proactive Improvement: When biases are found, sponsor banks must demonstrate efforts to identify and implement fairer decision-making systems. For partner banks with multiple lending programs under sponsorship, managing fair lending risks across many partners, many models, many data sources, and many decisions can seem daunting. That’s why our platform was purpose-built for sponsor banks: it provides a unified dashboard of fair lending risks and opportunities for remediation across many fintech partners. Our fairness-as-a-service tools: ➡ Consume an infinite amount of data ➡ Evaluate the fairness of the overall customer journey as well as each high-stakes decision made along the way; ➡ Monitor fairness outcomes and send alerts if fairness degrades; ➡ Generate less discriminatory alternative models; ➡ Quickly assess the fairness impacts of proposed changes to models and data sources; and ➡ Render results in easy-to-understand-and-explain visualizations. We also provide sponsor banks with the underlying data, computations and outputs of our system so that independent third parties can double check and reproduce our work. In today's regulatory environment, advanced fairness analytics aren’t optional – they’re essential. With fairness-as-a-service solutions like FairPlay, if you end up on the regulatory radar, you might just be doing it for the bragging rights!
