Richard Pace
Fool's Gold 4: The Brittleness of Algorithmic LDA Credit Models
This article examines the further risks and challenges of using less discriminatory alternative (LDA) credit models amid CFPB scrutiny.
top of page
The AI LendScape Blog
Search
Richard Pace
- May 21
- 38 min
Algorithmic Justice: What's Wrong With The Technologists' Credit Model Disparate Impact Framework
Current credit model disparate impact frameworks dispense justice algorithmically. But are they consistent with applicable fairness laws?
652 views
Richard Pace
- Mar 12
- 10 min
Navigating AI Transformation: Avoiding The Pitfalls of Algorithmic Sprawl
AI delivers rewards, but the C-suite/Board must address the risks of unchecked algorithmic sprawl diluting returns - governance is crucial.
106 views
Richard Pace
- Jan 30
- 20 min
The Road to Fairer AI Credit Models: Are We Heading in the Right Direction?
My research suggests that LDA credit models may pose risks to early adopters. What research priorities could put us on a more viable path?
308 views
Richard Pace
- Nov 25, 2023
- 30 min
Fool's Gold 3: Do LDA Credit Models Really Improve Fairness?
LDA Credit Models, believed to reduce credit barriers, may instead expand credit access inappropriately, thereby creating risks for lenders.
327 views
Richard Pace
- Oct 25, 2023
- 20 min
Fool's Gold 2: Is There Really a Low-Cost Accuracy-Fairness Trade-off?
The existence of similarly-predictive models with improved fairness is a foundational pillar of algorithmic debiasing. Is it actually true?
264 views
Richard Pace
- Aug 8, 2023
- 28 min
BISG Proxy Bias Redux: The Impact of the 2020 U.S. Census Data
How Does the Updated 2020 Census Data Impact the Accuracy of Race/Ethnicity Proxies and Fair Lending Disparity Estimates?
567 views
Richard Pace
- Jun 5, 2023
- 15 min
The Cross River Consent Order: Why Banks and Fintechs Should Be Concerned
The Order creates potentially serious conflicts on data and information sharing and fair lending compliance risk management responsibilities
763 views
Richard Pace
- Apr 22, 2023
- 13 min
The 10 Attributes of Highly-Effective Model Risk Managers
On the 12th anniversary of US regulatory guidance on model risk management, I discuss 10 attributes of highly-effective model risk managers.
1,271 views
Richard Pace
- Mar 29, 2023
- 21 min
Meta's Variance Reduction System: Is This the AI Fairness Solution We've Been Waiting For?
In this blog post, I review the promises and potential pitfalls of Meta's Variance Reduction System - a new tool to address algorithmic bias
552 views
Richard Pace
- Jan 27, 2023
- 36 min
Fool's Gold? Assessing the Case For Algorithmic Debiasing
Can we really achieve increased credit model fairness with minimal trade-offs?
950 views
Richard Pace
- Dec 14, 2022
- 10 min
Are Rising Interest Rates Impairing AI Credit Model Fairness?
How are macroeconomic conditions impacting AI credit model decisions and what does this imply for popular algorithmic de-biasing practices?
145 views
Richard Pace
- Sep 15, 2022
- 25 min
Using Explainable AI to Produce ECOA Adverse Action Reasons: What Are The Risks?
CFPB Circular 2022-03 raises questions about the use of Explainable AI to produce Adverse Action reasons. Here, I explore the key risks.
609 views
Richard Pace
- Aug 10, 2022
- 14 min
Algorithmic Bias and Alternative Data: Five Lessons From the DOJ's Meta Settlement
Where are federal regulators and enforcement agencies heading with their clear concerns over algorithmic bias in financial services?
486 views
Richard Pace
- Jul 1, 2022
- 13 min
Dark Skies Ahead: The CFPB's Brewing Algorithmic Storm
I examine the shifting winds in the CFPB's sentiment towards AI-based lending and describe two potential dangers brewing on the horizon.
615 views
Richard Pace
- Jun 10, 2022
- 15 min
Six Unanswered Fair Lending Questions Hindering AI Credit Model Adoption
In this post, I highlight 6 critical open questions related to AI bias for which consumer lenders need regulatory guidance.
662 views
Richard Pace
- May 16, 2022
- 17 min
Is the CFPB Adopting a New Fair Lending Proxy Methodology (BIFSG)?
What do we know about its properties, and are there important risks or limitations we should be evaluating now in anticipation of its use?
1,225 views
Richard Pace
- Mar 11, 2022
- 9 min
Debiasing AI Credit Models While Preserving Explainability
In this post, I do a deeper dive into the important interdependency of credit model explainability and fairness.
291 views
Richard Pace
- Feb 10, 2022
- 14 min
Four Potential Pitfalls of AI Credit Scoring Models
The last few years have seen rapid advances in the application of AI/ML technologies to the consumer credit underwriting process. In...
833 views
Richard Pace
- Jan 12, 2022
- 8 min
Six Fair Lending Focus Areas For 2022: What to Expect and How to Prepare
What's in store for lenders in 2022 as a new class of financial regulatory leaders ramp up their supervision and enforcement activities?
297 views
bottom of page