By Ray Birch
MADISON, Wis.—Is it “AI or die” among credit unions today? Two experts contend the movement has certainly reached a critical juncture and that credit unions—if they have not already—must begin strategically laying plans for use of the new data tools.
They also shared tips for doing just that.
“AI is not an if,” said Lamont Black, associate professor of finance at DePaul University and the leader of Filene’s new Center for the Credit Union of the Future. “With that in mind you have to treat it and give it the priority it needs, because it is coming your way whether you like it or not.”
Black shared that observation and more, along with John Best, CEO at Best Innovation Group, during a Filene Research Institute webinar titled “The Evolution and Impact of AI in Credit Unions.”
Best described artificial intelligence as a tool credit unions can apply to serve their members, and he urged CU leaders to think about new and innovative ways to provide service delivery.
“It is not necessarily a replacement,” Best said. “There are so many kinds of narratives around AI and robots replacing people. I do think there's a lot of fear in the market right now. But (in learning more about AI) you can get a little more comfortable with it and think about what you can actually do with it.”
The Strategic Approach
Black stressed credit unions must approach AI strategically.
“It's not just a product, it's not a service. It's not something you're just going to bolt on. You really need to form a strategy,” Black said.
Best said credit unions, as they consider AI and its potential uses, must be careful not to “de-risk themselves out of business.”
“We’ve got to find the sweet spot and the balance between AI and risk,” he said.
Best described how organizations have responded to developments in AI during 2022 and 2023 as being “like the gold rush.”
“Everybody had this amazing year where they got to experience ChatGPT for the first time,” he said, asking attendees to respond via chat if they had used ChatGPT in some form at their organization.
Lamont Black
As the chat function filled with responses, he added, “Wow, look at that. My goodness, that's nuts. So many people have used it and I know you've found incredible use-cases.”
Best pointed out that use of generative AI in business is gaining faster adoption than any previous technology.
“More than the telephone, the TV, the Internet…the list goes on,” he said. “C-suite executives are increasingly interacting with these AI tools directly.”
The Key Applications
What are the key AI applications being employed by organizations?
“I've seen people doing legal reviews, job descriptions, writing for marketing, product development, service operations and more,” Best said. “There's big value being delivered through generative AI.”
Black believes 2024 will see many more credit unions adopting AI.
“I am convinced that this year is going to be the year of AI, not for just the broad industry, but especially for credit unions,” he said.
During his presentation Black addressed all of the different forms of AI, such as generative AI, large language models, natural language processing and more.
Learning the Terms
“If you start learning about this, you'll learn many terms that are going to be new and unfamiliar. But gradually you'll start to appreciate what they mean,” Black said. “I would encourage you to start researching these terms and to then start to think about how can they be used; how they work. You don't have to get into all the details, but at least get comfortable with them and familiar with them, because then you can start to brainstorm ideas about what you could do with this technology.”
Black emphasized his belief that credit unions, like all organizations, are at a “tipping point” with AI, largely because it is easy to use.
“AI is user friendly,” he said. “Several years ago, AI was something that only a vendor did, and it was a black box. You didn't really understand it. So, you had to make a decision about whether you might want to use that or not. Just look at chat. Think about all the use cases. Start thinking about what it would look like to use some of these tools in the credit union.”
Chatting About ChatGPT
Black believes ChatGPT has created a national conversation around AI.
“Most of us have now used that. Some credit unions are using AI for purposes like market research, or entering a new geographic market,” Black said. “I think a lot of credit union boards are just now asking the question, what are we doing with AI, because it's in the news. Many of them are using ChatGPT, and that is now creating some pressure at the management level.”
For a credit union to finally decide it must begin deploying AI can be an “overwhelming” prospect, Black recognized.
Where to Begin?
“How do you begin?” he asked, before answering his own question and sharing some steps he advises credit unions take.
“Just simple kind of steps you could be considering as you walk down this path,” he said. “Get organized, start listening, empower your employees, access your knowledge, explore use cases, collaborate with others and create member value.
‘Not a One-Time Thing’
“As you think about an AI initiative, credit unions have hundreds of different initiatives…,” Black continued. “Every initiative has some champion and a committed group. So, you should be thinking about who's going to help lead this. Who's going to be involved in trying to implement this change? This cannot just be a bolt-on agenda item.”
As CUToday.info has previously reported, Black has spoken at numerous credit union events and with individual credit unions over the past several years as they explored cryptocurrencies.
“Sometimes it felt like just a curiosity. Like, ‘OK, let's show up. Let's have a conversation and then we can get back to work’,” he said. “AI in 2024 can no longer just be a curiosity. Don't just put it on the board planning agenda as a one-time thing just to scratch that itch.”
Processes for Working Together
Black encouraged credit unions to instead begin creating processes for working together as an organization.
“That might be a committee, and it might be something that's more informal,” he said. “But you should definitely be having these conversations regularly inside your credit union. You can start by just identifying and categorizing existing tools that you're already using that are very similar to AI. You can think of them as AI adjacent, like robotic process automation, or machine learning, which is part of AI.
“You might be surprised to find there are parts of your organization that likely are already using this. Maybe you already know that, and so you've got one step down the road and you can start thinking about how to build on the work that you've already been doing.”
Black encouraged credit unions to think of think of AI as the “new automation.”
“AI can do a lot more things than just automate,” he noted. “But, if you think about fintech and the last few years, a lot of it's been about automation. And AI is going to take this to the next level, so you don't have to start from scratch. You can build on things you've already been thinking about.”
The Next Step
Moving to the next step, Black stressed CUs must start listening, especially if staff are using ChatGPT. He noted that raises privacy issues, as ChatGPT will also learn from what the credit union inputs.
“This can be problematic because of the privacy and security issues around ChatGPT,” he said. “Many credit unions are starting to create a policy document around ChatGPT usage—what's good, what's bad, what's appropriate, what's not appropriate. I want to emphasize that should be a working policy document. You want to think about how do we put some guardrails around this, but also how do we also think about why people are using this.”
Forming New Groups
Black encouraged credit unions to form listening groups.
“Ask those users why are these tools helpful? Why are you applying this in this way? And that could be as simple as writing an email or some marketing content,” he said. “Clearly this tool is more than just about writing things. I would encourage you to iterate on that policy and learn from your employees. They are a source of information. Don't just treat them as a risk to be monitored with some policy document. Try to learn from them through these listening sessions…Don't just think of it as a risk management process. Think of it as an ideation session. How do we start to come up with ideas from the bottom up, not just from the top down.”
Getting a Copilot
He also encouraged credit unions to consider using Microsoft Copilot, a chatbot developed by Microsoft.
“Microsoft, very smartly, is now starting to incorporate that large language model (ChatGPT) into some of the services they offer,” Black said. “This is Microsoft embedding that large language model into the Microsoft 365 suite--Word, PowerPoint and Outlook. Think about a large language model as your words. How can your employees start to use natural language, in many cases just speaking English into one of these software designs, to think about how it can actually guide work using large language models instead of point and click drag and all of these sorts of traditional software interfaces.”