The Brave New World of AI Lending


When you think about AI, what comes to mind? A Hollywood blockbuster where robots become sentient and take over the world? While that’s definitely an entertaining possibility, the more likely scenario is that AI will have a much more subtle – albeit no less profound – impact on our day-to-day lives. 

One area where AI is already making its presence felt is in lending. Yes, lending – the process of giving someone money with the expectation of being paid back over time. AI has the potential to make lending faster, more accurate and reduce the costs of application processing. Let’s take a look at how AI is transforming lending and how it might benefit or harm borrowers in the process.

AI application in assessing creditworthiness

“Even people with high credit scores think being represented with a three-digit number doesn’t feel right,” says Dave Girouard, CEO at Upstart, a lending platform that works with credit unions and banks and offers personal loans. Girouard believes that AI provides unique opportunities to both lenders and borrowers because of its potential to refine borrowers’ credit profiles by including factors regular credit scores miss, such as education and employment. 

Upstart’s algorithms can quickly run vast amounts of data and offer the best interest rates and money management advice suited to fit each customer and their current situation. Around 70% of all credit decisions on this platform are instantaneous and fully automated.

Upstart claims it can offer better loan terms than traditional banks. Loan applicants with less than ideal FICO scores might be eligible for lower rates with companies that implement AI to assess the creditworthiness of their clients. That’s because, in addition to the standard criteria, such as the client’s FICO score, income, and credit report, the AI tool also considers details such as customers’ GPA, area of study, attended colleges, standardized test scores, and work history to gain better insight into each customer’s personal and financial ability to pay off a debt. 

This all may sound overly intrusive to some, but banks and other lenders seem to think highly of Upstart’s services: In 2021 alone, they paid $801 million in fees to the company. Upstart now plans to expand its successful line of businesses into mortgages (in 2023) and small business and auto loans.

AI-generated money management plans and advice

Another interesting example of AI use in the financial world comes from Israel, where a fintech company called Personetics Technologies has created a tool that assists banks in making more efficient, personalized money managing plans for their clients. 

Through its Engage tool, based on real-time adaptive technology that relies on various data points, Personetics examines customers’ transactions and analyzes their college graduation dates, transaction records, marital and parent info, and more. The results are sent to banks, which implement suggested changes, for example, offer better product rates or give personalized financial advice to customers. 

Is AI really unbiased?

Everyone can agree that lending in the US has come a long way since the times prospective borrowers had to meet in person with bank managers for an interview so that their creditworthiness could be assessed first-hand. Since the introduction of FICO scores in the 1950s, the whole process has been gradually evolving to become more transparent, accurate, faster, and less costly. 

AI lending is just the latest wave of innovation in that same direction. AI-enabled credit assessment platforms use machine learning algorithms to analyze borrowers’ credit history and other data points in order to predict their future behavior and the likelihood of defaulting. Some believe these tools can make lending more inclusive by neutralizing human lenders’ unconscious biases. However, many people forget that algorithms aren’t inherently more just than people. After all, they are just “recipes” that serve the agendas of those who made them.

In the context of lending, this could translate into decisions that are even more harmful to already marginalized borrowers. For example, an AI platform could recommend increasing the interest rate for a borrower from a low-income or minority group or denying them a loan altogether, based on data that shows they are more likely to default. In other words, increased accuracy will not serve all borrowers equally: It could simply reinforce the existing disparities in access to credit. The same goes for insurance, for example, where the Big Data revolution is threatening to entirely exclude some people as “uninsurable.”

We need to be aware of these dangers and cautious in our embrace of AI lending. As always, we should ensure that technology serves people – not the other way around.

Photo credit: The feature image has been done by Dusko Jovic.

This guest article has been submitted by Mila Bera. While we appreciate guest contributions, it's important to note that the views expressed by the author are not necessarily reflective of those held by TechAcute.

Was this post helpful?

Guest Author
Guest Author
This article has been submitted to us by an external contributor to TechAcute. We appreciate all external contributions but the opinions expressed by the author do not necessarily reflect the views of TechAcute.
- Advertisment -
- Advertisment -
- Advertisment -
- Advertisment -
- Advertisment -
- Advertisment -