AI Is Not Just Changing Financial Services. It Is Changing How Financial Institutions Are Built

Financial institutions have spent the last few years asking: Where can we use AI?

A better question is emerging:

What would a financial institution look like if it were designed around AI from the start?

That idea shaped Miami Fintech Club’s Executive Breakfast: Deploying AI in Financial Services, hosted in partnership with Blackbird on September 17 in Brickell.

Moderated by Alejandra Slatapolsky, Co-Founder of Miami Fintech Club, the discussion featured Driss Temsamani of Citigroup, Emilio Iturmendi of Microsoft, and Blake Rose of Addepar.

The conversation focused on what happens when AI moves beyond pilots and becomes part of the financial operating model.

Here are four ideas that stood out.

‍

1. Stop Adding AI. Start Redesigning the Process.

Temsamani discussed rethinking corporate account opening, a process traditionally filled with documentation, KYC/KYB requirements, reviews and operational handoffs.

“We didn’t want to do a use case or pilot. We wanted to build an operating system in the back.”

The shift is significant.

Instead of using AI to make employees slightly faster, institutions can rethink who does the work. Clients can initiate processes, AI can perform checks, and operations teams can focus on exceptions and oversight.

As Temsamani put it: “It’s an org chart change.”

That may be where the biggest AI gains emerge. Not from doing the same work faster, but from redesigning how work gets done.

2. Data Without Context Is Not Intelligence

Financial institutions already have enormous amounts of data across emails, CRM systems, calls, transactions and documents.

The problem is connecting it.

Rose captured this clearly:

“This recording means nothing. That recording has to be integrated in the taxonomy of your business.”

A transcript alone has limited value.

Connect it to the customer, account, opportunity and transaction history, and AI can begin answering much more valuable questions.

What did this customer ask for? What happened in previous conversations? What should happen next?

That is when AI stops being another application and becomes an intelligence layer for the organization.

3. You Don’t Have to Replace the Core to Create Value

Legacy technology remains one of banking’s biggest constraints. But institutions do not necessarily need to rebuild everything before improving the customer experience.

Iturmendi discussed a banking implementation where AI created a simpler experience on top of existing infrastructure.

The results he shared included an approximately 35% reduction in certain contact-center interactions and some credit approvals moving from as long as 48 hours to real time.

The lesson is simple:

Modernize the experience while you modernize the infrastructure.

AI can become a bridge between legacy complexity and better experiences for employees and customers.

4. Autonomy Does Not Eliminate Accountability

As AI agents take on more work, financial institutions face a critical question: Who is responsible when something goes wrong?

Temsamani’s answer was direct:

“You will never lose your responsibility.”

AI may check information, execute workflows and reduce manual intervention. But accountability remains with the institution and the humans responsible for the process.

For financial services, the future is not simply autonomous AI.

It is auditable autonomy, with clear controls, traceability, exception management and human accountability.

At Miami Fintech Club, we bring together the founders, financial institutions, operators and technology leaders building what comes next. Join the community and be part of the conversation.

‍