By Todd Michaud, CEO, HuLoop Automation
Discussions about AI and its usage in financial institutions have become nearly ubiquitous in financial media, if not among bank and credit union leaders. It is difficult to gauge how many banks and credit unions have begun implementing AI, but financial institutions are engaging with AI with cautious optimism. The caution paired with that optimism is certainly understandable. It can be challenging to know where to begin, especially as AI advocates purport – rightfully, I would argue – that this new technology can radically improve many credit union processes.
Beyond implementation challenges, other concerns are more fundamental. How will this AI impact day-to-day operations? Will it take anyone’s job? How will AI affect the member experience? Ultimately, these anxieties stem from a more essential concern about control. How can credit unions adopt this new, game-changing technology while keeping the human in the loop? This is the key question in AI adoption for credit unions. Without engaging with it thoughtfully, mass adoption of AI in credit unions will be impossible. However, suppose credit unions can utilize AI technology in ways that empower their employees. In that case, those credit unions will realize the full potential of AI to improve efficiency, drive growth and enhance the member experience.
An example will help illustrate this point. One of the most obvious and widespread use cases for AI in credit unions and many other industries is for customer service. Generative AI, including large language models, has demonstrated an impressive ability to process input and provide helpful outputs. This makes it an ideal tool for customer service, as it can take customer questions and either provide useful links or access and transmit the desired data. As implemented in credit unions, AI can accept questions from members about anything from the status of their loan application to information about new offers, and the AI can then search for the requested information on the back end.
Still, the question remains: How can credit unions use AI like this to increase efficiency without jeopardizing the member experience? The key is to give employees a way to intervene in automated processes. Returning to the example above, credit unions can – and should – utilize AI tools that allow employees to monitor communications between the chatbot and members and step in upon request or for any other reason. Thus, the credit union enjoys the benefits of enhanced efficiency, employees remain in control of communications, and members get the same or better member experience.
This principle extends to many other AI use cases in credit unions. Credit unions can utilize AI for automated underwriting while setting parameters within the institution’s risk tolerance. Credit unions can use AI to increase back-office efficiency by automating repetitive tasks, and employees can see all those automations, intervene in each process, and map their own automations to make their work easier.
In each of these cases, the AI tools keep the human in the loop. While improving efficiency, these tools are meant to improve the worker’s experience rather than replacing them. This also aids in implementation, as credit union workers will actually want to use their new tools instead of begrudgingly going along with the program. Only when credit unions find and implement AI tools that both enhance efficiency and improve the workers’ experience can they hope to utilize AI to its fullest. While AI technology continues to advance, the AI arms race in the financial services industry continues alongside it. Credit unions cannot afford to wait while their competitors invest in AI. Instead, they should seek out and implement AI tools that improve efficiency and keep humans in the loop.