🔒Bankers tread cautiously into an AI-assisted future seen as inevitable

At KeyBank, artificial intelligence is sometimes used to relay simple answers on customer service calls, but hasn’t yet been harnessed to vet borrowers or write drafts of shareholder reports.

KeyCorp Chairman and CEO Christopher M. Gorman said his company is taking a “significant” look at how to incorporate AI, but is moving cautiously, wary of potential biases in the technology and possible regulatory concerns.

So far, the $195 billion-asset bank, which has a major Connecticut presence, has only deployed AI in its call centers, Gorman said in a recent interview with the Hartford Business Journal.

“I think there will be significant applications for (artificial intelligence) and, ultimately, I think it will be good for workers because it will free them up from doing more rudimentary things, to do more value-added work and spend more time talking to clients and prospects,” Gorman said. “But it’s very early days for banking, and we’re not using it to a significant degree right now.”

Gorman is not alone in his assessment that AI will eventually be an important tool for the banking industry. But naturally risk-averse bankers in Connecticut and elsewhere are approaching the emerging technology cautiously.

On Naugatuck-based Ion Bank’s website, a chatbot named Fiona (Friendly Ion Assistant) can relay simple pre-programmed answers but can’t generate unique content on its own.

Ion President and CEO David Rotatori said his $2.2 billion-asset bank is unlikely to incorporate true AI — a machine thinking for itself — to a large degree anytime soon.

Rotatori said Ion is extremely cautious with its customers’ money and personal data and isn’t rushing to entrust either to a thinking machine.

“To interject things like AI, that are not yet known quantities, we are very risk averse, so we are going to take really small steps if and when we think AI is safe and applicable to what we are trying to do,” Rotatori said. “Allowing a machine or artificial intelligence to think for itself and make its own decisions that customers didn’t want, we aren’t going to take that chance.”

Ion does run a financial advisory firm that uses AI for research, Rotatori said. But any conclusions are vetted by professionals.

Rotatori said he hears similar sentiments from colleagues at other banks.

“Right now, everybody is very cautious,” he said.

Last year, Danbury-based Union Savings Bank, with $3 billion in assets, announced it was dipping its toe in the AI waters by launching an artificial intelligence-supported mobile app tool that allows customers to track spending and better manage their finances.

USB Spending Insights, which Union Savings launched in partnership with Florida-based financial technology provider FIS, offers more than 50 different data points about an individual’s finances and allows customers to track spending trends and set budgets.

It also analyzes accounts, predicts how much money users will have and suggests actions that can be taken to gain control of finances, the bank said.

Great deal of curiosity

Will Callender, president of banking technology provider FinMain, said every financial institution is likely using AI applications to some degree, whether they recognize it or not.

“I’d say all banks are using AI at some point today because all banks have vendors, and vendors use AI,” Callender said. “Lots of banks use analytic packages for pricing models, the price of a loan, price of a mortgage, price of a deposit. Generally speaking, those are outsourced solutions that rely on a lot of centralized intelligence from an analytic package that is processing data. That’s predictive AI, right?”

FinMain provides banking technology to dozens of midsize regional and community banks — ranging in size from $750 million to $150 billion in assets.

Callender acknowledged that the banking industry, on the whole, has not embraced the most modern forms of AI into everyday, in-house processes.

Even so, there is a great deal of curiosity from banks on how to leverage generative AI, which uses data to create original content. Callender said it’s one of the hottest topics at FinMain’s twice-yearly BankTank innovation forums.

There are a number of potential AI uses for banks, he said.

It holds the potential to cut down on fraud by recognizing suspicious activities and even flagging suspect checks. It can tailor marketing to individual customers and ease their banking experience.

It can also generate reports and eliminate much of the grunt-work involved with data entry. That could mean a reduced need for staff, Callender said. But it will also mean remaining staff will have more time to innovate, focus on customer service and build relationships.

“There is going to be a massive opportunity and responsibility to change the workforce from doing repetitive, simple manual tasks to more value-added tasks, which is no easy feat,” Callender said.

According to a June 2023 report published by consulting giant McKinsey, generative AI could have a significant impact on the banking industry, generating value from increased productivity of 2.8% to 4.7% of the industry’s annual revenues, or an additional $200 billion to $340 billion.

“On top of that impact, the use of generative AI tools could also enhance customer satisfaction, improve decision-making and employee experience, and decrease risks through better monitoring of fraud and risk,” the report said.

Early adopters in the banking industry have used generative AI primarily for software and knowledge applications, the report said. Other uses could include leveraging generative AI bots trained on proprietary knowledge — such as policies, research and customer interaction — to provide employees with round-the-clock, deep technical support.

“For example, Morgan Stanley is building an AI assistant …, with the aim of helping tens of thousands of wealth managers quickly find and synthesize answers from a massive internal knowledge base,” the report said. “The model combines search and content creation so wealth managers can find and tailor information for any client at any moment.”

Keeping up

James C. Smith, former CEO of Webster Bank and owner of banking consultancy JCSmith Advisors, said it’s sensible for heavily regulated bankers to tread cautiously with AI.

He expects banks deploying AI processes to duplicate those efforts with humans, at least short term, to double-check the technology’s accuracy and effectiveness.

This is the “belt-and-suspenders” approach banks used a decade ago when first adopting automated loan applications, Smith said.

Smith anticipates AI’s integration into banking will be subtle, but compounding and necessary for those who want to keep up.

“In the near-term, things don’t seem to be changing all that much, but over a five-year period they will change more than you can recognize,” Smith said. “I would say those bankers who don’t embrace AI as a tool that can benefit their customers, improve their marketing and communications, lower their operating costs, prevent fraud and elevate their employees, they’ll be challenged to keep up competitively.”