Ep 285: How MIT Leaders Are Building AI That Works w/ Michelle Jackson and Edgar Berrio (MIT)

This week, Michelle Jackson and Edgar Berrios, internal audit leaders at MIT, join the show. Drawing from their work at MIT, they share a practical perspective on how AI is being integrated into the audit process. Their approach goes beyond the technology itself, emphasizing the importance of guidelines, mindset, and human oversight to create AI solutions that are effective and responsible. The conversation explores how AI is transforming audit teams today. From defining agentic AI to practical use cases like drafting audit findings and improving communication, Michelle and Edgar break down what’s actually working in practice. They also discuss the importance of clear governance and human review to ensure quality outputs. The episode also includes a real-world walkthrough showing how to build an AI agent using Copilot. Be sure to connect with Michelle and Edgar on LinkedIn. 1:32 - Best Chat prompts 5:10 - Agentic AI Explained 7:00 - Prioritizing Company Use Cases for Agentic AI 10:20 - Developing Guidelines and Guardrails for Agentic AI 14:23 - How to Build an Agent 25:55 - Drafting an Audit Finding 32:17 - Final Thoughts   Be sure to connect with Mark on LinkedIn. Also, be sure to follow us on our social media accounts on LinkedIn, Instagram, and TikTok.   Also be sure to sign up for The Audit Podcast newsletter and to check the full video interview on The Audit Podcast YouTube channel.   This podcast is brought to you by Greenskies Analytics, the services firm that helps auditors leap-frog up the analytics maturity model. Their approach for launching audit analytics programs with a series of proven quick-win analytics will guarantee the results worthy of the analytics hype.    Whether your audit team needs a data strategy, methodology, governance, literacy, or anything else related to audit and analytics, schedule.