By Soumoshree Mukherjee
Editor’s note: This article is based on insights from a podcast series. The views expressed in the podcast reflect the speakers’ perspectives and do not necessarily represent those of this publication. Readers are encouraged to explore the full podcast for additional context.
The hum of AI is getting louder, promising transformations across every industry. But amidst the excitement, a crucial question often gets overlooked: how do you actually implement AI successfully, especially in complex, regulated environments like financial services? It turns out the answer isn’t just about sophisticated algorithms or cutting-edge tech. It’s about people, strategy, and governance.
In an illuminating conversation on “The RegulatingAI Podcast” between host Sanjay Puri and Jeff McMillan, head of Firmwide AI at Morgan Stanley, McMillan challenges the tech-first mindset and lays out a compelling blueprint for responsible AI innovation.
READ: Dr. Girish Nadkarni’s global vision for future of AI in healthcare (
Forget the dazzling tech demos for a moment. McMillan makes a bold assertion: the biggest hurdles to AI adoption aren’t found in the code, but within the organization itself. “The reason people are failing is not because of their understanding of the underlying technology,” he states. “They’re failing for the same reasons they’ve been failing since the beginning of digital transformation: a lack of organizational focus, a lack of leadership, a lack of engagement, clarity of strategy….” It’s a powerful reminder that without clear leadership, strategic alignment, and a solid connection between technology and business objectives, even the most advanced AI tools will flounder.
Morgan Stanley wasn’t just observing the AI revolution; they were helping to lead it, partnering with OpenAI as early as March 2022, well before ChatGPT became a household name. This head start allowed them to craft their own “rules of the road” in the absence of established regulations. Their governance strategy rests on four pillars:
- Senior Supervision: Ensuring full transparency and oversight from the highest levels.
- Human in the Loop: Acknowledging AI’s helpfulness, but firmly placing ultimate responsibility and accountability with the human. As McMillan aptly puts it, “the humans press the button.”
- Robust Evaluation: Recognizing that while Generative AI is easy to build, its non-deterministic nature makes high-efficacy deployment challenging. This demands rigorous, both automated and manual, input/output testing frameworks.
- Second-Line Oversight: Establishing an independent group to validate every system before it goes live, adding a crucial layer of control.
READ: Imperfect data, real impact: Inside Abhishek Mittal’s AI philosophy (
One might assume a regulated entity would move cautiously, but McMillan argues there’s a “risk of moving too slow.” AI isn’t just a productivity tool; it can improve a firm’s risk posture by enhancing monitoring and reducing output deviation. Furthermore, to mitigate bias, Morgan Stanley feeds its own curated knowledge into LLMs and rigorously tests models against demographic populations to ensure fairness.
Perhaps the most critical takeaway for any organization embarking on an AI journey is the paramount importance of employee education. Investing in training on what AI is, how to craft effective prompts, and guidelines for responsible use isn’t a luxury; it’s a necessity. It empowers the workforce, fosters responsible adoption, and builds an AI-literate culture from the ground up.
Drawing on his background at West Point and in the U.S. Army, McMillan shares three invaluable leadership lessons: Be structured yet flexible, empower your teams rather than centralizing power, and always maintain perspective.
McMillan’s insights paint a clear picture: successful enterprise AI isn’t about magical algorithms, but about thoughtful strategy, robust governance, a deep commitment to human oversight, and continuous education. As Puri wisely concluded on the podcast, “Responsible AI isn’t a constraint on innovation, it’s the foundation for… sustainable and trustworthy innovation.” It’s a lesson every organization venturing into the AI frontier would do well to heed.