Rory Brown, Managing Partner of Nicklaus Brown & Co., Shares How AI and Machine Learning Can Improve the Virtual Bank Experience
The banking industry is undergoing a period of deep disruption. Virtual banks have reimagined what banking can do for customers, and AI and machine learning represent a new area where online-only banks can dominate traditional banks.
These two technologies represent the vanguard of digital banking, but because traditional banks are operating on legacy systems, it’s difficult for them to fully integrate these advancements. Virtual banks, on the other hand, are relative newcomers, operating within newer systems that aren’t hindered in the same way. This means virtual banks can move more quickly to provide their customers with the benefits that AI and machine learning can bring.
Humans are notoriously bad at accurately judging risk. It’s the reason we fear getting on a plane, even though planes are immensely safe, while not thinking twice about hopping behind the wheel of our cars, which harm far more people every year.
Machines don’t share the biases that skew our judgment. Plus they’re much better at finding patterns in large datasets. Banks can use machine learning and AI to analyze vast volumes of historical customer data to find correlations between certain behavior patterns and adverse outcomes. When these profiles are applied in real-time to customers, banks can receive warning signs long before customers default on mortgages or make other financial mistakes.
This is also useful when determining credit risk. These technologies are much better at determining the creditworthiness of an individual than human operators. They can look at more data points, more quickly, and create a thorough portrait of a customer’s likelihood of getting themselves in trouble. This makes it faster and easier for customers to qualify for loans and saves banks from potentially poor investments.
Machine learning algorithms can use historical customer data to build extremely accurate models of normal spending behaviors, and it can do this in real-time for every customer. When anomalous spending patterns are detected, AI can quickly assess the level of threat and then determine a course of action. For low threats, it might require an additional layer of authentication. For more severe threats, accounts can be locked out immediately.
This level of fraud monitoring is entirely impossible for human operators, as the datasets requiring analysis are far too large and far too complicated.
AI and machine learning allow banks to offer new customer service options that don’t require paid human labor. Chatbots are a great example of this technology. AI and machine learning technologies create smart chatbots that use natural language processing to understand what customers are asking, and then deep transaction and behavioral analysis to provide novel answers to difficult questions. This level of personalization is normally reserved for human operators, but smart chatbots will bring that service for pennies on the dollar.