NEW YORK — Financial institutions are diving headfirst into gen AI but University of Michigan Credit Union hasn’t seen much in terms of its return on investment.

Amid financial institution hype around gen AI, “the ROI is not that great,” Sherry Wu, chief technology officer at University of Michigan Credit Union, said at FinovateFall 2025 Sept. 8.

The $1.3 billion credit union has deployed AI-driven chatbots for consumer-facing uses and internal applications, Wu said.

Calculating ROI for internal uses can be difficult as it can show up as “employee morale increase,” Wu said. “If the contact center agents are freeing up their time because AI chatbot is able to answer simple questions, then they are able to answer more complex questions, more strategic thinking and build that relationship with the customers.”

READ MORE: ROI of AI: Is the juice worth the squeeze?

The Ann Arbor, Mich.-based credit union is working with AI services provider  interface.ai and rolled out its AI-driven services in July, according to the fintech’s website.

UMCU has deployed multiple gen AI use cases with interface.ai, including:

• AI-driven chatbot “Vic”; and
• AI assistants for customer services teams.

The AI-driven customer servicing tools are available around the clock while traditional business hours end at 6 p.m., Wu said.

She did not disclose specific AI spend, but the credit union reported a noninterest expense, which includes tech spend, of $43.4 million in its 2023 annual report, up 16% year over year.

Looking beyond low-hanging fruit

According to the “GenAI Divide: State of AI in Business 2025” report published in July by MIT, only 5% of AI pilot programs achieve rapid revenue acceleration while others stall or show no impact on the bottom line.

FIs need to look beyond customer servicing to find ROI on gen AI investments, Wu said, adding that the use of agentic AI can drastically help FIs to make operations efficient and generate revenue.

Check out our exclusive new bank industry data here.