AI in Banking: Top Priorities for 2022 (And Beyond)

Banking executives know they need to integrate AI into their institutions' daily operations. The potential use cases are almost limitless. Some, like chatbots, are well known, but not yet widely implemented. Here are five of the most promising applications of artificial intelligence in banking.

Artificial Intelligence in financial services is still largely filled with untapped potential. While many bankers may think of things like chatbots and fraud monitoring when it comes to AI, in reality the technology can be used in just about any conceivable part of a bank or credit union.

For the most part the industry has not even scratched the surface of how AI can transform banking. Here are some of the top ways financial institutions can deploy artificial intelligence in 2022 and beyond.

Smarter Chatbots and Digital Assistants

Let’s start with the most obvious application first. Some banks and credit unions already deploy chatbots, but as of now most perform only basic functions, with some exceptions. The ones that use more advanced artificial intelligence to offer predictive insights are more typically referred to as digital or virtual assistants. But as AI and natural language processing become more advanced, expect more of these intelligent digital assistants to play a bigger role in consumer’s financial lives.

Several surveys and market research studies have found that people actually prefer interacting with bots instead of humans. Although some of these surveys are conducted by conversational AI vendors, skeptics should consider the fact that there are 24 million users of Bank of America’s digital assistant Erica and they completed 123 million interactions in the fourth quarter of 2021, up 247% year over year.

Beyond Simple Bots:

Conversational AI is enabling chatbots to offer predictive and personalized financial insights to customers, a trend that will only continue.

“The quality of chatbots will definitely improve over the next few years,” states Light IT, a software solutions firm. “They will predict human behavior more accurately and use this information for self-learning.”

USAA, for example, upgraded its Eva virtual assistant in 2022 to understand more customer “intents,” including when they use slang.

Chatbots also save time and money by reducing human interactions with customers. Juniper Research forecasts that chatbot interactions will save 862 million hours for banks globally, which equates to $7.3 billion in cost savings. That’s just one area AI can help create efficiency, as covered next.

Creating Internal Efficiencies

If you have been in banking for even a short while, you are likely already sick of hearing the word “silos.” Nevertheless, it is indeed true that many financial institutions have a siloed structure, and often the systems and technology used by different lines of business do not “talk” to one another — at least not in ways that can be readily used. This leads to internal inefficiencies and unnecessary manual work.

As noted by McKinsey in an AI report, most banks’ data is fragmented across separate business and technology teams, and analytics efforts are focused narrowly on stand-alone use cases. This makes it harder to get a full view of each customer, and offer personalized, customized digital services.

AI can help connect that data and enable banks to perform “analysis from internal and external sources at scale for millions of customers, in (near) real time, at the point of decision across the organization,” McKinsey notes.

This ultimately will not only create internal efficiencies, but facilitate a greater customer experience.

Banking customers transact through ATMs, physical branches, contact centers and mobile and online banking — creating friction in the journey, observes James Freeze, chief marketing officer for AI company Interactions, in Forbes. “Consumers shouldn’t have to go searching for insights into their mortgage, savings or investments. These insights should be proactively served up across channels, and AI can facilitate that.”

Read More: The Future of Customer Experience in Banking is Personalized

A More Personalized Customer Journey

It’s well known that customers want more personalization and customization from their banks. But many institutions fall short of delivering such an experience. AI can help in this regard.

AI-powered data analytics can enable banks and credit unions to better understand customer needs, says Jim Marous, Co-Publisher of The Financial Brand and CEO of The Digital Banking Report. This will ultimately enable them to act as a “concierge” to customers, proactively providing insight to them based on real-time financial opportunities or threats.

“The power of data, advanced analytics and artificial intelligence will be at the foundation of consumer engagement, creating autonomous, real-time decisions without human intervention,” Marous states in a white paper.

Golden Data:

One of AI's biggest potentials is to help banks convert their 'treasure trove' of data into useful insights for customers, and wallet share for the institution.

Major retail companies like Amazon, Netflix, Kroger and PayPal already tailor product discounts and recommendations to their customer base using data analytics and artificial intelligence, notes financial education technology company Everfi in a blog. AI can help financial institutions dig into their data to find similar opportunities.

“With no shortage of customer data, financial institutions are sitting on a treasure trove of answers in terms of where customers are headed next and what their financial needs will soon be,” Everfi writes. “Armed with that data, institutions can grow wallet share and generate revenue by catering their products and services to customers in anticipation of their time of need.”

Webinar
REGISTER FOR THIS FREE WEBINAR
How Modern is Your Core? How FIs Can Start Their Digitization Journey
In this webinar, attendees will learn real-world examples of how banks took a phased approach to start their digital journey and the ROI of implementing a modern core.
Thursday, April 11th AT 2:00 PM (ET)
Enter your email address

Lending and Credit Decisions

Artificial Intelligence can play a vital role in not only making credit decision faster, but ensuring that banks are easily able to identify applicants with a high probability of paying back the loan.

Creditworthiness of an individual, until now, has largely been based on past credit history and current earnings. AI, on the other hand, can power predictive models of a person’s ability to repay a loan, as opposed to relying on historical data.

Expanding the Loan Pool:

Combining alternative data and AI enables banks and credit unions to make better lending decisions and lend to a wider range of borrowers.

By using AI to analyze data, banks can qualify new customers for credit services,
determine loan limits and pricing, and even reduce the risk of fraudulent loan applications, McKinsey states.

“Setting themselves apart from traditional banks … AI-first banks have designed streamlined lending journeys, using extensive automation and near-real-time analysis of customer data,” the firm notes.

AI can also be used to support more inclusive lending, by enabling access to a broader range of financial products, streamlining the application process to make it more accessible and performing credit scoring free from human bias.

“For lenders, it combines a socially oriented strategy with a great marketing tool,” observes Dmitry Dolgorukov, Co-Founder of HES FinTech, in a BAI post. “Borrowers, even those with thin files, get a better chance to receive financing at a reasonable price. For the financial industry as a whole, expanded access to finance without compromising the financial stability of lenders and borrowers looks like a great match.”

One caveat: So-called “black box” lending has also drawn scrutiny from members of Congress and the Consumer Financial Protection Bureau. Financial institutions need to be able to demonstrate that the AI technology they use doesn’t incorporate built-in biases.

Read More:

Faster Response to Fraud and Cyber Security Issues

With the sheer number of transactions that banks must process today, it’s incredibly easy for fraud to slip through the cracks. In just one area, P2P payments, scams are rising exponentially. Fraudsters are using increasingly sophisticated tools and tactics to bypass traditional defenses designed to detect and stop fraud.

Stunningly, the ratings agency Fitch has said that only about 1% of money laundering activity currently gets detected. It’s no surprise then that fraud and security is one of the top use cases for artificial intelligence that bank executives are looking at, according to an Economist Intelligence Unit survey.

Banks are looking at AI not only to reduce losses and more efficiently use resources, but to improve customer experience too, the survey says. “Mastercard, for instance, uses data on transactions and authorizations to predict and detect fraud more precisely and quickly: reducing false positives means fewer legitimate transactions are stopped, improving customer experience.”

It’s not just fraud like money laundering, but attacks that directly target customers that banks must be wary of. In 2021 banks experienced a 70% increase in account takeover attacks targeting bank customers, according to Kevin Gosschalk, CEO of fraud and security company Arkose Labs.

AI can be used to help detect and stop these attacks.

“Quickly evolving attack threats make it hard for fraud teams at banks to get out ahead of bad actors,” Gosschalk tells The Financial Brand. “That is why AI is an absolutely critical component to effectively detect and stop fraud.”

For example, banks can use AI to better understand online traffic patterns, Gosschalk adds. The technology can help determine valid device fingerprints specific to customers over a given time period, which helps banks determine the right action and response to take in near real-time.

This article was originally published on . All content © 2024 by The Financial Brand and may not be reproduced by any means without permission.