Your Next Loan May Be Based More on Your Smartphone and Other Data Than Your Credit Score
Image: Pixabay

Your Next Loan May Be Based More on Your Smartphone and Other Data Than Your Credit Score

Artificial intelligence is integrated into many aspects of business, including financial risk analysis, and can predict your loan repayment behavior

March 31, 2020

The days of using a handshake to secure a loan are a distant memory. Today, artificial intelligence powers all sorts of business practices, from what Walmart sends to each store to which boxes should go on delivery trucks. But artificial intelligence goes beyond shipping. In fact, your next loan decision could be based upon an artificial intelligence as much as a credit score. So how does it work?

How Does it work?

When looking at Artificial Intelligence (AI), we see two different kinds: supervised and unsupervised. Supervised AI involves the input of certain rules by humans into the software, which allows the AI to learn based upon those rules or values. Think about how your phone is able to tell the difference between face photos or how web searches can recognize a what's in a photo if you upload it into the search. People had to teach the AI how to distinguish these things, after which people should only have to make tweaks to ensure accuracy. A lot of larger financial institutions use supervised AI in order to make lending decisions. The software uses the things people put into the system, such as underwriting rules, collateral values, payment history, credit history, etc., to make a credit worthiness decision.

Unsupervised AI is often used by smaller institutions, especially for borrowers with limited to no credit history. Instead of feeding the software predetermined variables, the AI instead analyzes all available data on borrowers in order to look for patterns. In this case, it matters more how you fit into the data groups as opposed to what kind of credit score you have. The AI might determine that some groups of borrowers might default on loans at a much higher rate, such as those who file insurance claims or those who post on social media late at night. And if you fall into these groups, you could find yourself facing heavy scrutiny.

AI reduces overhead

When we look at AI, we might think that it's all bad news or that lenders are getting too involved in our data. But when larger lenders use AI to assist with the loan process, it can reduce overhead and streamline the process, thereby reducing delays and getting things wrapped up quicker. Some companies even use AI to to automate the entire loan process, which could lead to reduced bias, better loan terms, and quicker (or instant) decisions.

Beneficial for limited credit histories

If you have no credit history, what are your odds of securing a high-dollar vehicle loan at a good rate? The chances typically aren't good without a co-signer. But what if unsupervised AI could use those various data factors to predict that you would be a good borrower? This could help new borrowers who fit into the data sets the lender likes to see qualify for loans they would otherwise not be able to get. Just like it might make lenders scrutinize you if you have a lot of negative tenancies, you could find yourself qualifying for the better loans and better rates because your data suggests you'd be a good borrower. While we don't know what those factors are, we can come up with some made-up possibilities to give you an idea. If your social media posts are limited, if your phone is set to do not disturb while driving, or if you pay your utility bills early as opposed to on the due date, you might signal that you will be a good borrower.

Applying with apps can send data

You might notice that some lenders ask you to download an app as part of the application process, possibly in order to streamline the process and make it easier for you. But you need to be mindful that downloading an app could expose vast amounts of your data to the lender, such as daily location history or patterns, text message punctuation, whether your contact lists has last names, and even what times of day your phone is actively in use. This data is sometimes called "alternative data" and is basically your digital footprint consisting of more than 10,000 different variables. It is often used to make decisions about short-term, high-interest loans.

Fraud Prevention

AI has the ability to prevent loan fraud, which cuts loses for the lenders and should translate into better loans and rates for the borrowers. AI has been used for years to prevent credit card fraud, which has grown exponentially in recent years due to the rise of e-commerce. The AI can analyze a customer's buying habits and location in order to trigger a fraud alert. Similarly, financial institutions can use AI to analyze patterns of credit application, as well as money transfer in order to detect money laundering.

Data Privacy Concerns

When AI is put to good use, the potential benefits can be great. It can help people be more productive, better predict risks, and help to remove bias. But the use of AI, whether in the financial sector or elsewhere, is still relatively new and isn't fully understood or fully implemented. Because of this, it will be evolving still for some time to come. But with the relative newness come concerns over privacy, which many can argue comes from the intrusion of these companies into our personal data points and using them to gives us a score. Besides using this data to make decisions about credit worthiness, it's really impossible to know where the data is going after it is used. Time and time again, security breaches have demonstrated that data sometimes gets stored, shared or even sold without your permission, even if the collecting company has a strict data privacy policy.