Committee decisions, data silos, and COBOL: what's slowing AI in banking

As Principal and Partner leading Next-Gen Banking and Payments Tech at EY, Mitch Siegel spends his days with CIOs, Chief Digital Officers, Chief Architects, and heads of retail and commercial banking - the full committee of people trying to move banks forward in the age of AI.

In this episode of Banking Reinvented, Mitch joins host Tim Rutten to make the case that the AI gap in banking is neither a model problem nor a compute problem. Instead, it is an architecture problem, a data problem, and an operating model problem. A study EY co-led with MIT Technology Review across 250 banks put a number on it: only 16% have agentic AI in production, and even that figure, Mitch suggests, might be generous.

The conversation covers why AI transformation in banks is governed by committee and why that is precisely what slows it down, how the regulatory framework banks have treated as a drag is about to flip into a competitive advantage, and what workforce displacement due to AI looks like in organizations. 

Mitch also walks through how EY itself is reshaping from a pyramid to a diamond structure as AI absorbs the junior work and the premium shifts to people who can navigate complexity at the top.

-

Curious to find out more about AI in banking? Check out our content hub.