Insight

Explainability and Artificial Intelligence

March 07, 2024

EQUIFAX HAS DRIVEN RESPONSIBLE ARTIFICIAL INTELLIGENCE (AI) innovation for nearly a decade. We use AI Systems in a transparent, trustworthy, fair, explainable, and secure manner, to provide benefits to consumers and customers. 

In 2015, we led the way towards an industry standard for Explainable AI (xAI), introducing the first machine learning credit scoring system with the ability to generate logical and actionable reason codes for consumers. 

Why is this so important nearly a decade later? Explainability remains one of the central commitments of Equifax AI. It is important in ensuring that the decisions assisted by AI are fair, ethical and understandable. Ultimately, explainability provides accountability in the use of AI, providing detail on what informed a decision.

Equifax leverages AI to help solve complex problems and to introduce new methods of modeling that can ultimately help more people access mainstream financial opportunities and drive greater financial inclusion. One example of this is our OneScore consumer credit model – the first offering in the industry to combine traditional credit history with telecommunications, pay TV and utility payment data on over 191 million consumers, as well as Equifax DataX and Teletrack specialty finance data on 80 million consumers - including payment history from non-traditional banks and lenders.

Through these alternative data sources, OneScore delivers a more robust financial picture of consumers who are actively seeking new financial services, including credit cards, personal loans and automotive loans. Powered by advanced analytics and machine learning, OneScore can potentially increase credit scores by up to 25 points and the scorable population by more than 20%, or 8.8 million credit-seeking consumers, when compared to traditional scoring models.

We leverage AI techniques in model development when Equifax research indicates that the inclusion of these advanced techniques will increase the predictive outcome of the model. More than 70% of our credit scoring models around the globe were built by Equifax with AI in 2023 – and we’re not done innovating.

We led the way in xAI and now, as of March 2024, we have more than 90 approved AI patents supporting our approach to AI, primarily for decisioning and fraud prevention. We also have more than 130 AI patents pending as we continuously work to harness the power of AI and advanced data science to help people live their financial best.

For more information about AI at Equifax, please visit Equifax.ai.