Despite all the buzz, generative artificial intelligence won’t change banking. But it’s poised to change how banking is delivered.
That’s according to Michael Abbott, global banking lead at consulting firm Accenture. Banks seeking to capitalize on generative AI’s potential are increasingly focused on the technology handling time-consuming tasks so employees’ time is freed up for revenue-generating efforts, such as more face time with clients, he said.
Interest in generative AI has ballooned over the past 18 months, and even the risk-averse and highly regulated banking industry has scrambled to explore potential uses. Early adopters of genAI could see up to a 30% productivity improvement over the next three years, according to an Accenture report published this year.
Banking could be one of the industries most affected by the technology, and top Wall Street banks are making more moves to incorporate genAI into their operations. Last month, Goldman Sachs rolled out its genAI tool for code generation firm-wide, increasing developers’ efficiency, The Wall Street Journal reported. And this month, Morgan Stanley is launching a tool called Debrief, which maintains detailed logs of wealth advisers’ meetings with clients and summarizes discussions, according to CNBC.
Generative AI is already embedded within much of the technology that banks buy, or employed by vendors they work with, including legal firms, Abbott said. And many banks are now using genAI to handle call summarization at call centers.
But financial institutions are pursuing other uses, too, including tapping the technology to develop scripts for marketing purposes, creating synthetic identities banks can treat almost like focus groups, or finding red flags in mortgage applications, allowing loan officers to spend more time with clients, Abbott said.
Even more promising, Abbott said, is genAI’s ability to translate legacy code, which could benefit a host of banks trying to modernize old systems.
“Generative AI is impacting every single level of the bank, but the biggest impact may be actually turning it on technology itself and acting as a Rosetta Stone, to take us out of the 50 years of legacy and bring us into a completely modernized world in a very efficient way,” Abbott told Banking Dive in a recent interview.
Editor’s note: This interview has been edited for clarity and brevity.
BANKING DIVE: How are banks of various sizes approaching generative AI use?
MICHAEL ABBOTT: It’s not just the big banks. When it comes to the more sophisticated stuff, if you're going to use this for coding, you obviously have to have your own code. So some smaller community banks are just going to buy capabilities. But we've seen small credit unions using it for marketing and things like that. It’s impacting everything from the front-end — marketing, conversations, call center — all the way through to the back end.
Because a lot of it’s cloud-based technology, it’s being democratized, so even the smallest bank in the world can take advantage of it.
There’s no doubt generative AI can generate efficiency. The biggest opportunity, though, is on the revenue side. How can you take waste out of the processes and systems, but then put the value back in? The very best banks in the world are thinking about it that way. The revenue side opportunity far exceeds the cost side. Let's say I've got a loan officer in a commercial banking space who’s writing credit memos all the time. If I can automate the writing of that credit memo, I can have that loan officer out there talking to more people.
Publicly, everyone’s talking productivity; privately, the best banks are talking revenue. They’re thinking, I’m using this to free up my people's time to focus on revenue generation.
Where do you see banks remaining cautious or even hesitant when it comes to using genAI?
Certainly, there's a variation in the take-up rate. It's not correlated to size; it’s more correlated to culture. The banks that have a culture of curiosity tempered with execution — meaning they're biased toward action, not just thinking about things — anybody that has that culture is kind of in this space. There are a bunch of other banks that I would say are more dismissive, and don’t have that culture of curiosity, and those are being held back.
Still, I don't know of any bank that has turned generative AI over to its customers. You augment with genAI; you automate with AI. AI is a very predictive model. AI takes a set of variables, you run it through the process, you get an output, you can test, predict, you can ensure risk and compliance. You can explain it completely to the regulators. Generative AI, depending on how you set what’s called the temperature of it, meaning the variability that you're willing to let it have, will generate different answers, given the same input, at different times. Most banks are augmenting with generative AI, using it to augment human processes, and still having a human in the loop, but not completely automating it and putting it out in front of the consumers.
What’s yet to come with genAI use?
I think it's back to the future in banking. Think about what we have to do today to open a mortgage, now that it’s digital. We have to go to some website, enter our information, upload documents, go through that “journey” that that bank created for us. Fifty years ago, we would have walked into a branch, have somebody help us get the mortgage, process it and take care of us.
Now, because of generative AI and AI, we can start a conversation, chatting with our banker on our mobile phone, where we can have a virtual conversation, we can just upload documents. There's no more app, no more website, no more journey to be pushed out. Instead, we just have a conversation, and they’ll walk us through and tell us where we are and what we need to do.
Oddly enough, in becoming digital, banks have become functionally correct and emotionally devoid. The big opportunity for technology is to put humanity back into the experiences again.
