Industry expert provided thoughts on how credit unions can best leverage artificial intelligence (AI).
By John San Filippo
The second annual Reseda Summit, hosted by the Reseda Group, commenced Sept. 18 and concludes Sept. 19, 2024. The Reseda Group is a wholly owned CUSO of MSU Federal Credit Union that invests in fintechs beneficial to the credit union movement. The Summit’s purpose is to showcase more than two dozen Reseda portfolio companies for credit union executives from around the country.
With 2024 dubbed the “year of AI” by the University of Texas at Austin, among others, it’s not surprising that the Summit kicked off with a discussion about AI. The keynote speech was delivered by Henry Iversen, co-founder of boost.ai, who flew in from his native Norway to address the audience. The company provides AI-powered chatbots and other self-service technology to financial institutions and other organizations around the world. After sharing the recipe for “Norwegian moonshine” (50% moonshine and 50% coffee), Iversen got down to business.
“I'm not a credit union expert, but obviously I've been working a lot with special financial services firms in terms of AI as a field,” Iversen told the audience. “That is something I have big passion for in my company.” He added that he founded boost.ai in 2016 and counts MSU Federal Credit Union among the company’s many longtime clients.
Getting Started
Iversen noted that while mega banks and neobanks lead in terms of dollars invested in AI, credit unions have access to AI tools to level the playing field. The challenge, he added, is staying current on what’s going on in the AI space.
“I actually had my first summer vacation this year,” said Iversen. “I was away four whole weeks. And just within that timeframe, there were a lot of things going on in the marketplace. [In the past] when you summarized what's going on in the AI space, you could maybe do that in five minutes throughout the week. Now you need maybe 30 minutes. Things are moving extremely fast and the question now is: Are we able to adapt?” He added that the institutions that can adapt fastest to changing technology are the ones that will succeed going forward.
Iversen then presented data from a Gartner Group study that surveyed more than 100 financial institutions on AI adoption and implementation. “There's actually 30% that have not done anything,” he observed. “That’s surprising to me. And then we have another 29%, which are planning to do something; and then we have another 29% actually developing pilots right now.”
The message to credit unions, according to Iversen, is that they need to act now on AI or risk falling behind. “It's not too late,” he said. “There's still time, but there's not a lot of time. We definitely need to move on this.”
On ChatGPT and Generative AI
Iversen told the audience that while generative AI technologies like ChatGPT have been transformative, financial institutions may have overreacted. “ChatGPT coming out was definitely a massive shift for us,” he explained. “But in the early days, you could see organizations having a budget for ChatGPT technology or ChatGPT projects or having a strategy around ChatGPT, which I don't think is necessarily helpful. We need to have an AI strategy, but it's not necessarily only about the ChatGPT technology.”
He continued: “When you communicate [with ChatGPT], it seems supersmart. You can ask questions, and you get pretty reasonable responses back. But the problem is that you are not able to develop and build your strategy based on this technology because it's only based on the data we have given it. If we give it a task which doesn't have a lot of training data, it's going to struggle.” He warned the audience that there are “gaps you need to work with.”
The Challenges
Moving on to the challenges credit unions face in developing an AI strategy, Iversen said, “The first one is around change management and communication. I think this is a big challenge because in many cases, it’s about automating some employees’ tasks, automating part of their job. We work a lot with the public sector and it's sometimes pretty hard to get people in the public sector to automate their jobs because that means that they also will potentially need to do something else in the organization.” He noted that effective communication is critical. “How are we going to speak about this internally?”
AI is also likely to face increased regulatory scrutiny, Iversen noted. He pointed to the EU AI Act, a regulatory framework proposed in 2021 and implemented in 2023, as an example of what may be coming.
“We need to be ready to adapt to new regulations,” he said. “We need to document things. We need to have governance around how we deliver AI because I'm pretty sure there's going to be similar things in the U.S. I think there's also a lot of internal work already in this area.”
Organizational Buy-in
According to Iversen, for an AI strategy to be successful, it needs to be embraced by the entire credit union. “Please, please don't outsource the AI strategy to the IT department,” he said. “It’s super crucial that as an organization, you need to work with AI technology.”
He continued, “If you're launching something within finance or HR, make sure those people are also accountable for that project. Again, it’s not just an IT project. We need that whole organization accountable.”
Comments