Fair lending refers to the principle and practice of ensuring that all individuals and businesses have equitable access to credit and financial services, free from discrimination based on race, color, religion, national origin, sex, marital status, age, disability, or other protected characteristics.

Many think the fair lending is the end state we should all strive for. But we think the logic in that goal is flawed.

Best explained by SOLO founder, Georgina Merhom in her own words…

On a recent panel at American Fintech Council (https://www.linkedin.com/company/american-fintech-council/), I was asked how AI could make lending less biased.

My breakthrough on this topic didn’t come from my experience at SOLO serving banks. Ironically, it stemmed from my early career in public policy and, ultimately, why I decided to leave it.

The full-circle moment came to me during a conversation with Alex Johnson (as usual, our side conversations could double as podcasts). We discussed the myth of “unbiased reporting” in journalism and how the same concept applies in public policy, and in credit.

Paraphrasing Alex: “Would you rather I openly share my bias so, as the reader, you can account for it as context? Or should I pretend I have no bias, no specific lens or context, and ask you to trust me regardless?”

The same principle applies to credit. The idea that a single, black-box credit assessment can be “unbiased” is mathematically unviable. Every assessment is a lens, and that lens assigns weights to specific attributes, filtered through a particular context.

Personally, I believe credit shouldn’t strive to be “unbiased” or “inclusive” in the way these terms are often discussed. Credit, by design, is exclusive and inherently biased toward individuals who meet certain requirements within a defined context.

To me, claiming an assessment is unbiased is naive at best and an illusion of fairness at worst. Instead, what we should strive for — especially when discussing fair lending — is transparency.

Would you rather trust a black box that claims to be “unbiased” but offers no insight into how it reaches decisions? Or a transparent framework where the weights, assumptions, and logic are clearly laid out — enabling you to understand the criteria, provide additional context, and even adjust your financial behavior to achieve your desired outcome?

In a transparent, collaborative data sharing environment, you can clearly see what factors are being considered and how they are weighted. This transparency empowers actionable steps, moving beyond a one-time “fair” assessment to an ongoing partnership between a bank and its customer towards a long-term financial roadmap in an evolving context.

Ironically that’s what best-in-class relationship banking was like before automation was ever introduced.