Day 2 at Fiserv Forum: Who Owns the Intelligence?
Day 2 at Fiserv Forum: Who Owns the Intelligence?
By Jeff Bassill
Posted August 19, 2026 | Category: Industry Watch
Today's sessions at Fiserv Forum shifted away from company-specific updates and toward the future of the financial services industry as a whole. Unsurprisingly, artificial intelligence touched nearly every session, whether the topic was member service, operational efficiency, cybersecurity, analytics, compliance, or software development.
As with most industry conferences, I left with as many questions as answers. Two concerns stood out. The first is our tendency to overgeneralize what we mean by “AI.” The second is whether we're adequately considering the long-term implications of relying on it.
Artificial intelligence has become the catch-all phrase of 2026, but not all AI is the same. Some of what gets called AI is a large language model, generating text and human-like responses through a process that's largely a black box to the user. Other AI is closer to an advanced decision engine, weighing factors and adjusting outcomes based on historical patterns, more sophisticated than traditional automation but still fundamentally different from generative AI. When we talk about AI's risks and benefits, it matters which one we're actually discussing. Treating fundamentally different tools as if they carry the same opportunities and challenges is how good risk conversations go sideways.
Human in the Middle, or Human in Charge?
One message that repeatedly surfaced from Fiserv leadership was that AI must be both secure and transparent, concepts that matter enormously in an industry where trust is one of our most valuable assets. More than once, speakers emphasized keeping a “human in the middle” of the process.
I agree with the principle, but futurist Amy Webb added a distinction I think is critical: it isn't enough to have a human in the middle. We need a competent, knowledgeable human in the middle. That raises a bigger question. Is the human's role really to sit in the middle of the machine, reviewing its output? Or should the human be leading the machine?
I prefer the latter. “Human in the middle” pictures a person as another cog in a complex set of gears. “Human in charge” says the technology is a tool directed by human judgment, expertise, and accountability. That distinction matters.
AI and the Disappearing Learning Process
A recent conversation with my son brought this into focus. He earned his Master's Degree in High Energy Nuclear Physics in 2018, and the coursework behind it was punishing: pages of handwritten mathematical proofs, complex programming assignments, extensive data analysis, none of it graded merely on the correct final answer. The point was understanding the process required to get there. The learning happened while solving the problem, not after.
Today, AI could complete a meaningful share of those same assignments. That raises a real question: what knowledge and analytical skill will a student graduating ten years from now actually possess, if the technology performs much of the work that used to build those skills?
The concern isn't that AI can solve the problem. It's whether future professionals will know how to solve it when AI is unavailable, wrong, or operating outside its intended boundaries. The same question applies to our industry. If AI drafts policies, analyzes regulatory requirements, reviews contracts, or summarizes compliance obligations, who develops the expertise to catch the errors, omissions, or hallucinations? The value of human oversight depends entirely on the capability of the human doing the overseeing.
Transparency Includes Cost
Fiserv leaders repeatedly emphasized transparency in how AI systems reach a decision, and I agree completely. But transparency should extend to the long-term cost of using that system, too.
Many emerging AI products could be transformative, especially for smaller credit unions that struggle to attract specialized talent or maintain large operational teams. The value proposition is real. But the pricing model has to be sustainable and transparent as well, priced on the technology's long-term economic reality rather than promotional pricing or loss-leader strategies built to drive adoption.
Current AI subscription pricing reminds me of rideshare a decade ago. Uber and Lyft transformed transportation with rides that felt remarkably cheap. The disruption was real, but a lot of those prices never reflected the long-term economics of the service. Today's AI market raises the same questions. If a technology is going to reshape staffing, operations, and strategic planning, organizations need confidence that the economics behind it are built to last.
Final Thoughts
The most valuable lesson from today wasn't that AI is coming. That's already obvious. The more important question is who retains the intelligence.
Bill Handel of Raddon made a similar point in a breakout session on economic cycles: the future doesn't belong to AI alone, and it doesn't belong to institutions that reject it either. It belongs to the ones that combine human and artificial intelligence, using each where it adds the most value.
Financial services should absolutely adopt tools that improve efficiency, sharpen decision-making, and expand what we can do. But we can't let those tools replace the human knowledge required to evaluate their output. Technology should amplify expertise, not eliminate the need for it. As credit unions evaluate AI over the next several years, I believe the institutions that succeed won't be the ones with the most advanced technology. They'll be the ones that keep investing in the human intelligence needed to use it wisely.
The views expressed here are my own and do not represent the official position of Kings Federal Credit Union.
