Beyond the AI Hype: Building the Credit Unions of Tomorrow

Artificial intelligence (AI) is transforming industries, and credit unions are no exception. Moving beyond the hype, strategically adopting AI is essential for credit unions to enhance operations. This webinar, moderated by Lisa Arthur, Sensedia's SVP of US operations and global CMO, features insights from industry experts Megan Tjernagel, CEO of Premier Credit Union, George Estrada, Strategic Advisor at AWS, Paul Dionne, CSO at Quantum Governance, and Paolo Malinverno, CSO at Sensedia, and explores how credit unions can navigate their AI journey effectively.

Full webinar available to watch below.

‍

Key Challenges for AI Adoption in Credit Unions

At FORUM Credit Union in Indiana, COO Andy Mattingly shared that their AI-driven underwriting process has led to a 70% increase in loan processing capacity. These are real results from AI, but implementing AI isn't without its hurdles. 

According to Tjernagel, credit unions face several significant challenges. One major  challenge, stakeholder resistance, leads to staff resistance and results in a lack of full commitment to AI adoption. She shares that Integration proves difficult, particularly when integrating new AI applications into existing, outdated core systems. Outdated core providers and vendors may limit AI functionality, creating obstacles. Beyond technical issues, concerns about the accuracy and safety of AI-generated information remain significant. Tjernagel also stresses the sheer scale of AI adoption leads to project fatigue. Teams know this is a huge undertaking.

Malinverno expresses that the entire banking industry is changing, albeit slowly, in large part because of consumers' wishes. Credit unions are seeking ways to provide those services.

Estrada points out that, like the ability to understand a company's finances, all c-suite leaders, beyond just the CTO/CIO, need a strong understanding of the organization's and industry's technology. It's no longer acceptable to say, "I'm not good with technology." 

Capturing the full value proposition of AI becomes difficult when relying solely on traditional ROI metrics. Estrada uses the example of a call center, using AI to cut call times from seven minutes to three. He asks, since consumers are willing to give you seven minutes of their time, should the objection be to cut down the call time or consider using the extra four minutes to cross-sell or meet other member needs? 

"We have to start rethinking everything," he shares. "If you're stuck only on evaluating efficiencies and ROI, you're missing the big picture, and somebody's going to get there before you."  

‍

Prioritize AI Adoption Strategically at the Board and CEO Level

Leadership and board understanding are absolutely critical for successful AI adoption. Dionne explains that AI is the future and says, "We cannot imagine credit unions surviving long into the future if they don't adopt AI in an effective way. It's part of the game now. It's where the future is going to be determined. Boards need to be part of that conversation."

He warns against the temptation for boards to sign off and let staff take care of things. The C-suite needs to educate boards about how adopting AI will benefit the credit union and its members.

Tjernagel explains that developing a comprehensive AI strategy cannot be limited to the IT team or CEO alone; it requires involvement from the entire C-suite and building a comprehensive strategy with leadership and the board. It's not OK to keep "kicking the can down the road." She shares, "Start small and figure out what needs to be done in order to get to the finish line." She continues, "Credit unions - they've got to start small; start simple. But they've got to get started." 

Dionne offers some practical steps credit union leaders can take to get started. He says, "The last thing you want to do is have your board be a bottleneck." He continues, "Let's bring them around. Let's make sure they're informed and understand the strategic plan so they can be promoting and pushing for that for you as a CEO, as a CIO, as a technology team." Bring the right information to the board to raise their awareness. Help them understand the risks and opportunities for AI adoption so they understand the strategic focus and path forward.  

Dionne discusses that there isn't a silo in the credit union that isn't already or won't soon be touched by AI. A comprehensive overview of the strategic plan is essential, including guardrails.  

‍

Regulating AI

The US is behind in regulation. Dionne says people are already using AI, particularly on their phones. It's up to the credit union and board to assess risks, ensure compliance, and approach safety and soundness internally. He believes the regulatory pieces will probably come after the fact, so credit unions need to lead on their own.

George Estrada compares AI's early stages to Social Media's. Initially, everyone thought social media was great. Today, we see the divisions and issues it has caused for young people. He suggests more business leaders and societies need to weigh in rather than leaving regulatory decisions solely to AI developers and governments. For example, as tasks become more automated, we could choose to ensure automation leads to more human-to-human contact. He says we must ask what things we will not do and have an ongoing, dynamic dialogue about what we do want to achieve. And it will be different from society to society. Estrada's opinions were highlighted in his article: AI Regulatory Framework: A Thought Experiment .

Dionne expands, "The board is the conscience of the credit union. They're the shepherds of the mission of the credit union." He explains the important role they play in thinking about where AI can take the credit union. AI is a different paradigm from other innovations. For example, without regulatory oversight, credit unions need to determine their path and, in doing so, understand that "only AI can effectively monitor AI." So, we'll use one form of AI as an oversight to another type of AI. Considering that AI learns and continuously changes, setting up compliance won't be as simple as setting it up and moving on because the thing that monitors compliance is changing.

Dionne expounds that beyond being the compliance guardians and keeping the credit union safe, the board needs to determine what types of risks to take to ensure the future of the credit union. Credit unions are, by nature, risk-taking organizations. In the ever-changing AI playing field, the board and leadership team need to have the conversation about the risks and how the organization will play within a constantly changing field.

‍

The Role of Tech Providers

Maliverno shares, "AI is not going to evolve according to what is good for people, I'm afraid. AI is going to be evolved by providers of AI technology that are driven by profit."  He brings up ethics and states, "We tend to sleepwalk into technology and find out the consequences later. We gave mobile phones to our kids, and now we wonder why they won't talk to us."  The bad actors are quicker at developing technology, so responsible leaders need to focus on ethics and guarding the use of  AI for good.

Malinverno adds that a lot of tech that brands itself is not really AI. He jokes about AI-enabled toothbrushes and shares that real AI has to make autonomous decisions and continue to evolve and learn.

Lisa Arthur brings up the F5 Report and shares that 96% of organizations are now deploying AI models, emphasizing the growing reliance on AI for agile, secure and scalable infrastructures to support these technologies. The report states that organizations need to focus on simplification and standardization of operations, including streamlining APIs, technologies and tasks. The report also highlights that 94% of organizations are deploying applications across multiple environments, including private clouds, public clouds and on-premise data centers. She asks the panel what steps credit unions should take as they move forward with AI technology.

Malinverno discusses that AI is good at some things and bad at others. For example, AI might help to process a loan application faster, but when denying a loan, it doesn't understand the consequences the denial has for the applicant who really needs it. Implementation takes time, trial and error and improvements to training AI so it can learn better. We take our time and move forward slowly, always have a human review of how we're training AI models because they are making decisions that impact real people.

Estrada shares that smaller credit unions will need a strong partner strategy. They'll be dependent on partners to move forward.  Partners are more than vendors, so ensuring they align with the credit union's missions and values is essential. Mid-sized credit unions will need to determine what partners they'll need and what they can keep internally. The biggest credit unions move away from partnerships, determining what stays in-house and where it makes sense to outsource.

He warns about taking it slow, "AI can scale and scale fast. And it's so fast that the damage is done before you notice that there's damage." Trial and error help teams determine which processes might need a human to be involved. While speed is everything, Estrada stresses to "hurry slowly."

API-First Strategy

Malinverno explains that many credit unions worry that using APIs to open and connect to other platforms will erode members. He then tells the panel that, actually, credit unions will lose members if they don't open their data. His international experience has shown that open finance allows for more opportunities for growth rather than mass exodus. And to experience open finance, APIs are the gate. They allow credit unions to open, connect and scale with more security.

How Will AI Improve Member Experience

Tjernagel expects AI to improve fraud detection as fraud continues to grow. She also believes AI will help her credit union understand member preferences, patterns and behaviors so they'll feel understood, which will build trust and loyalty. And she says, "Today, people want what they want. When they ask for something, they want it. I feel like we're going to be able to provide rapid responses to them with fast and consistent information." She is excited about the opportunities and says we may have underestimated what AI holds for the future of credit unions.  

Real AI Use Cases

AI is already being implemented in various practical ways within credit unions. Dionne discusses sentiment analysis for call center conversations to help staff de-escalate situations and get a better idea of the member's mood. He also likes internal chatbots for staff that contain all the internal documents and processes of the credit union to provide answers on policies and other internal questions. Tjernagel shares that her credit union is already benefitting from an internal system, PAL (Premiere Assistance Library), that provides this information and has increased efficiencies.

She also discusses how her credit union used AI to help with processing the due diligence of a merger. She explains they were careful about how they inputted information so as not to break any non-disclosure or confidentiality agreements. Her team was able to create time frames based on the targeted merger date and add state and NCUA requirements. They were able to compare policies, rates, fees, and exam reports. Overall, using AI simplified the process and created a map for future mergers.

What's Next for AI and Credit Unions

Estrada lights up when he imagines how AI might be used to analyze and evaluate behaviors, like emails and other member interactions, to help everyone in the organization align with the credit union's values. But ultimately, he says, most benefits of new technology don't appear until a decade or more. He cites GPS and how it started out as an expensive, less reliable tool, and now we have enhanced, extremely accurate navigation integrated into our phones and watches.

Navigating the AI journey requires a thoughtful approach. A recommended strategy is to be a "fast and savvy follower." This involves paying attention to what's working elsewhere in the industry, adopting it quickly, and then improving upon the adopted systems and tailoring them to the specific needs and goals of the organization. Credit unions should start small, build a roadmap, and progress toward their goals.

Final Words of Advice

Dionne agrees that it’s crucial to be a fast, savvy follower of AI adoption. Fast followers pay attention to what's going on and working in the market. Pick it up and bring it to your market. But be savvy as well. The best followers also improve on the systems they adopt and think about flaws and what could be better. They tailor solutions to their goals and what is best for their organization.

Malinverno shares that teams should experiment with AI and invite people on the team who don't know technology to join in the experimentation process. Your team should keep their feet on the ground, keep values and guidelines in mind, and stay focused on the good you're trying to do. Give AI the safe range to make autonomous decisions and request it refer to you for anything outside the range. This will help teams move fast, slowly. Use the head and the heart. Stay focused on ethics - something that is very hard to train an AI model to do.

Tjernagel shared to trust and verify. She also suggests credit unions recruit technology subject matter experts to the board, stating it's been very helpful for her organization.

Estrada wraps up the conversation, saying to "apply skepticism and humility. Keep that balance." He tells the panel that the people who are hyping AI the most have the least humility around it. He believes AI is one of the most awesome technologies out there - comparable to nuclear power. Be skeptical of people who claim to know what they're doing and how AI will impact civilization. Come to AI with some humility and start applying the ethics.  He talks about Frankenstein's monster, "The creature wasn't bad. What happened was he (Dr. Frankenstein) abandoned the creature to the world, and it learned and behaved the way it was treated. So this awesome technology, if you throw it out there and let anybody play with it, you're going to get what you put out there. And this is where the humility part comes in. We should know that there's a lot we don't know. So let's teach it what we want it to be, as best we want it to be. Be skeptical of the hype and bring a lot of humility into it, and I think you could make some really good decisions."

‍

Make sure to catch our upcoming webinars, where Sensedia brings together industry experts to share actionable insights, spark meaningful dialogue, and drive the future of API driven innovation.

‍

—

To explore more on Credit Union innovation, visit our dedicated page:

‍https://www.sensedia.com/report/credit-unions-into-future-success

‍

‍

‍

‍