U.S. Bank's Chief AI Officer on Strategy, Governance, and Scaling AI

Please support our episode sponsor:

Explore executive education programs from Emeritus, in collaboration with top universities.

======

Prashant Mehrotra leads AI strategy at the fifth-largest commercial bank in the US. In this conversation, he explains how U.S. Bank decides which AI projects to move forward, what it takes to scale AI in a regulated environment, how the bank builds AI that customers trust, and what must be in place before AI can act autonomously on a customer’s behalf.

You’ll learn:

The framework US Bank uses to evaluate and prioritize AI initiativesHow governance can accelerate, rather than slow, AI deploymentThe design principles that separate helpful personalization from customer intrusionWhat capabilities must exist before “do it for me” banking becomes realKey TakeawaysAI Transforms Processes, Not Just Efficiency

U.S. Bank's Chief AI Officer, Prashant Mehrotra, frames AI as an "intelligence layer" rather than a tool. 

The bank asks a fundamental question before deploying AI: Is this process necessary in its current form? This mindset shift moves teams beyond incremental improvements toward reimagining entire workflows. 

For example, the bank's generative AI developer assistant works as a "wingman" alongside partner firms, not simply answering questions but actively troubleshooting and collaborating. 

Leaders should challenge their organizations to question existing processes rather than settle for automating the status quo.

Governance Accelerates When Risk Partners Engage Early

Many organizations treat risk review as a gate at the end of AI development. U.S. Bank flipped this model by embedding risk partners from the start, cutting approval times in half over six months. 

The bank builds learnings from each AI deployment into its platform, making subsequent approvals faster and more repeatable. This collaborative approach treats governance as a feature of the AI platform itself rather than an appendage. 

Organizations stalling on AI deployment should examine whether their governance model creates friction or enables speed through early, continuous partnership.

Baselines Determine Whether AI Pilots Scale or Fail

Mehrotra points to the oft-cited statistic that 95% of AI pilots fail and attributes much of this to poor measurement discipline. 

The bank establishes baselines for current performance before launching any pilot, measuring not only speed but quality of outcomes. When AI-assisted code reviews exceeded expectations by more than 50%, the data enabled confident scaling to thousands of developers. Contact center response times dropped from minutes to tens of seconds with clear before-and-after metrics. 

Leaders must resist the temptation to launch pilots without rigorous baseline measurements, or they will lack the evidence needed to justify enterprise-wide investment.

Episode Participants

Prashant Mehrotra is EVP and Chief AI Officer at U.S. Bank, where he leads the organization's AI strategy and implementation. He is an accomplished AI executive, patent holder, and speaker, with extensive experience creating, building, and leading AI/ML initiatives across finance, insurance, and retail sectors. Prior to joining U.S. Bank, he was Head of the AI Center of Excellence at Allstate. Prashant previously held key leadership roles at Capital One and Staples, where he successfully implemented advanced analytics ecosystems and data strategies.

Michael Krigsman is a globally recognized analyst, strategic advisor, and industry commentator known for his deep business transformation, innovation, and leadership expertise. He has presented at industry events worldwide and written extensively on the reasons for IT failures. His work has been referenced in the media over 1,000 times and in more than 50 books and journal articles; his commentary on technology trends and business strategy reaches a global audience.