Data and Analytics

How AI is Transforming Traditional Credit Scoring & Lending

How AI is Transforming Traditional Credit Scoring & Lending

May 07, 2024 | Dave Sojka

Artificial Intelligence (AI) is all the rage right now. At a recent conference that I attended, a guest speaker was asked, “What’s the first move that you would recommend for companies just getting started with AI?” The speaker quipped, “Add an AI statement to your investor relations web page!”

Not bad advice, given the current trajectory of AI technology investments. Goldman Sachs projects global investment in AI to approach $200 billion by 2025.¹

The outsized investments represent big bets on the future. But the future of AI seems to evoke more questions, and emotions, than most innovations. Which industries will be most affected? Will AI create more jobs or replace human workers? What are the practical applications and limitations of AI?

Equifax has been driving responsible AI innovation for nearly a decade. We led the way toward an industry standard for explainable AI — introducing the first machine learning credit scoring system with the ability to generate logical and actionable reason codes for the consumer. The Equifax Cloud was custom built to manage the large volume of diverse, proprietary datasets needed to maximize AI performance and deliver AI-infused products.

Below are excerpts from a recent panel at Fintech Meetup where Harald Schneider, Global Chief Data and Analytics Officer for Equifax, shared his thoughts on how AI is transforming credit scoring and lending for Fintechs. 

Question: What types of AI models have traditionally been used by bureaus and lenders?

Harald: Historically, models used by bureaus and lenders have focused on traditional credit data and logistic regression models. Over time, there has been significant progress in both data collection and algorithm development.

Additional data sources, such as cash flow data and alternative lending data, have been integrated to supplement traditional data included on credit reports. This expansion in data sources has allowed for more comprehensive assessments of credit risk behaviors. Furthermore, algorithmic advancements have shifted from simplistic regression models to more sophisticated techniques like gradient boosted machines and neural networks. Regulatory compliance and explainability remain critical considerations in model development.

Question: How does AI apply to the credit ecosystem and fair lending practices?

Harald: A crucial aspect of credit risk modeling involves ensuring fairness and equity. Those of us who are focused on the credit ecosystem understand the importance of interrogating our models to ensure they treat all populations fairly.

We recently released our Responsible AI Policies and Principles designed to ensure consistency in best practices across different markets.

Question: Are there market or policy opportunities and impediments in the AI space?

Harald: As AI and data usage continue to expand in the financial industry, there are both opportunities and challenges to consider. Generative AI, which produces new outputs, holds potential benefits for various applications, including data processing, and consumer education.

However, it's important to note that generative AI may not be suitable for core credit risk models due to the need for factual explanations. While AI technologies offer exciting possibilities, it's crucial to navigate them responsibly, considering factors like ensuring regulatory compliance, fairness, and transparency.

Question: What advice do you have for fintechs on deploying AI effectively?

Harald: Fintech companies have a significant opportunity to leverage AI to enhance their operations and offerings. Mature applications in software development, quality assurance testing, contact centers, and vision AI have demonstrated proven return on investment. These areas offer tangible benefits for fintechs looking to incorporate AI into their operations. 

It's essential for fintechs to recognize that AI should augment human work rather than replace it entirely. By leveraging AI technologies in areas where they are most mature and well-understood, fintech companies can optimize efficiency and improve outcomes for their customers.

To learn more about AI at Equifax, visit Equifax.AI

* The opinions, estimates and forecasts presented herein are for general information use only. This material is based upon information that we consider to be reliable, but we do not represent that it is accurate or complete. No person should consider distribution of this material as making any representation or warranty with respect to such material and should not rely upon it as such. Equifax does not assume any liability for any loss that may result from the reliance by any person upon any such information or opinions. Such information and opinions are subject to change without notice. The opinions, estimates, forecasts, and other views published herein represent the views of the presenters as of the date indicated and do not necessarily represent the views of Equifax or its management.



Dave Sojka

Dave Sojka

Risk Advisor

Dave Sojka has over 25 years of experience in Consumer Credit Risk working at institutions such as Household International, CitiCards, Alliant Credit Union, and Check into Cash. He also spent time as an analytics consultant at TransUnion. During his time at Equifax as a Risk Advisor, Dave led the development of the Ris[...]