AI’s Future Potential in the Lending Industry

Alyssa Huang -

In recent times, Artificial Intelligence (AI) has come to play an increasingly important role within the lending industry. Before these technologies, an employee would have to manually review each loan application before deciding which ones to approve or decline. Now, AI softwares have started to do the same, and banks are relying on them more frequently, especially for large amounts of loan applicants. Already, some loan processes have become entirely automated. Furthermore, some AIs scrape not just the data available in the loan application but also data available online using social media sites such as Facebook to learn more about the loan applicant and determine their trustworthiness.

With the role of AI rapidly increasing in the lending industry, it is important to recognize that  AI — and technology in general — is not impartial. Rather, AI is an extension of and reflects the biases of those who built these systems. However, if designed in an ethical manner, AI has the potential to be both more unbiased and accurate than human judgement.

AI has already proven to have many benefits to the lending industry. Banks in particular benefit the most from this technology. Since AI can unearth trends indiscernible to humans, the banks have had an easier time in more accurately predicting a customer’s behavior or tendencies to default based on past behavior, outperforming employees in many situations. In this regard, banks will have a more accurate sense of how likely each customer is to pay back their loans, and they can minimize losses in this way. Additionally, AI can help banks attract new customers by taking data from their search history.

Although AI does provide many benefits to banks, the current lack of regulation in this domain raises many ethical concerns. In particular, anti-discriminatory laws only protect customers against discriminatory lending practices done by bank employees and have not yet incorporated AI into the picture. However, AI still has the potential to discriminate. At present, race impacts loan application. As a result, if an AI is deciding whether or not to accept a loan application using data from which past loan applications have been accepted or rejected, the software is unknowingly factoring human biases about race into its decision-making process and will output a biased decision. AI’s lack of transparency also brings about ethical concerns. When denied a loan application, applicants need to be informed the reason for their rejection so that they know what to expect with their next loan application. However, some AIs leave no indication of their decision-making process and rather only display their final decision (accept or reject), a practice which violates the rights of the loan applicant.

With the onset of Covid-19, experts have indicated that technology will play an increasingly large role in helping banks recover financially from the pandemic. As a result, many banks will likely turn to AI to help determine borrowers’ risks, as it is a quicker and seemingly more accurate option. Zest AI, a financial service company, predicts that even after the pandemic has ended, lenders will likely still continue to increasingly rely on AI to predict risk scores.

If implemented correctly, AI offers a unique opportunity to render the lending industry more fair and efficient. Although current systems lack the governance needed to ensure an ethical AI, if either given more unbiased data or a way to rectify its own mistakes and biases through a feedback loop, AI has the potential to be a great asset to the lending industry.

Sources

  1. “COVID-19 Has Changed Credit Modeling Forever. Here's How.” Zest AI, May 6, 2020. https://www.zest.ai/insights/covid-19-has-changed-credit-modeling-forever-heres-how.
  2. COVID-19 Impact on Lending Industry: Digital Growth to Combat Crisis. Accessed March 29, 2021. https://www.tcs.com/how-covid-19-will-affect-digital-lending.
  3. Goel, Nikhil. “AI in Digital Lending.” Wipro. Accessed March 29, 2021. https://www.wipro.com/blogs/nikhil-goel/ai-in-digital-lending/.
  4. Klein, Aaron D. “Fair-Lending Laws Haven't Caught up to AI.” American Banker. American Banker, May 2, 2019. https://www.americanbanker.com/opinion/fair-lending-laws-havent-caught-up-to-ai.
  5. Levin, Jeff. “Council Post: Three Ways AI Will Impact The Lending Industry.” Forbes. Forbes Magazine, November 5, 2019. https://www.forbes.com/sites/forbesrealestatecouncil/2019/10/30/three-ways-ai-will-impact-the-lending-industry/?sh=3758e3346899.