“AI’s disruption of banking is inevitable–for better or worse.’ So claims a recent column in American Banker. It’s a bold, maybe even alarming statement. But what does it really mean?

Artificial intelligence is a branch of computer science aimed at making computers think more like people. In particular, AI aims to make computers learn, so that the more information they take in, the better they do at recognizing meaningful patterns in that information. This process – which can be automated in a constantly improving feedback loop – is sometimes called machine learning. A computer algorithm takes in a huge amount of data, analyzes it, and identifies patterns, which then help it analyze and predict data even more accurately in the future.

The applications for banking are enormous. Enriched with AI capabilities, computers can take on many everyday tasks now done by human employees. For example, Swedenbank has a digital assistant that “talks” to customers 30,000 times a month. This chatbot resolves 78% of customer problems without any need for a human to help out. AI-powered computers can automate the more mundane, time-consuming part of an employee’s job, allowing them to focus on questions requiring higher-touch handling. The result? Customers enjoy better, more personalized customer service, while employees can devote themselves more fully to the interesting aspects of their work.

But the potential of AI banking algorithms goes way beyond answering simple customer questions like “why was I charged a fee?” Already, AI programs are empowering computers to do things like verify customer identities, do a first pass on a loan application according to a bank’s customized credit memo, and more. The Mirador platform is built from the ground up with powerful AI and machine learning capabilities. We can use machine learning to analyze customer data and get a better picture of creditworthiness, allowing us to recommend extending loans to small business owners who, for example, have average personal credit but very strong sales. Our AI programs also analyze repayment data, flagging loans that might be in danger of default, so loan officers can work with borrowers early to bypass problems. (Another great read on this subject: our previous post 4 Ways Machine Learning Makes Small Business Lending More Profitable.)

Banks that incorporate AI’s potential can free up enormous amounts of staff time. For example, loan officers spend 30 or even 40% of their time reviewing loans that are obvious no’s. The rise of AI means that computers can now do this simple work, freeing up staff to spend their time building relationships with strong borrowers, instead of wasting their time rejecting unsuitable candidates for loans.

Which bank do you think will be stronger and more profitable–the one whose staff spends their time manually checking customers’ IDs, looking up names in government databases, rejecting completely unsuitable loan candidates, and dealing with sudden, unexpected defaults? Or the one whose staff is equipped with algorithms to run simple checks and flag potential problems, and spends their time resolving complex issues, talking to customers, and building relationships in the community?

Artificial intelligence and machine learning are here to stay. The only question is, is your bank ready to take advantage of these new technologies?

Leave a Reply

Your email address will not be published.