Webinar Recap: Copilot to Coworker: Transform Identity Governance with Agentic AI

Identity Governance and Administration (IGA) is no longer just about compliance – it's a cornerstone of security, agility, and operational efficiency. In our latest webinar, Lumos CEO Andrej Safundzic sat down with Kevin Converse, VP of Identity and Access Management at GuidePoint Security, to unpack why AI is becoming essential in the IGA world, what challenges teams face today, and how to navigate this shift intelligently.

In the webinar, Kevin and Andrej cover:

• Why IGA Is More Critical Than Ever
• Two Big Challenges in IGA Today
• How AI Can Help
• Pitfalls to Avoid
• AI as a Layer, Not a Replacement

Let’s dive in.

Why IGA Is More Critical Than Ever

“IGA is not just a compliance requirement, but it’s really like the core of your security program,” Kevin explained. As companies grow, adopt cloud tools, and become more dynamic, managing who has access to what—and for how long—has become a top priority.

Andrej put it plainly: 

“What IGA does right is making sure that the right people have the right access for the right amount of time.”

With identity-related breaches on the rise, and most hacks involving compromised credentials, the urgency to get IGA right has never been greater.

Two Big Challenges in IGA Today

The conversation highlighted two persistent pain points:

1. Visibility and Integration: “You need to integrate with every app...and go down to the entitlement,” said Andrej. Kevin echoed that sentiment, noting how difficult it is for companies to gain real-time insights into application access and entitlement-level data: “Legacy tools were point-in-time. You really need to understand what's actually happening in those environments.”
2. Achieving Least Privilege at Scale: Creating meaningful RBAC/ABAC models is still a manual, error-prone process. “It’s a year- or two-year-long process, and you never get there,” Andrej emphasized.

How AI Can Help

The second half of the discussion focused on AI’s potential to bridge these gaps, without replacing the human element.

Andrej likened the current state of identity to a “stick shift” car, saying, “We don't need to be fully autonomous yet…The question is, how can we get to driver assistance?”

Kevin highlighted where AI can help most:

• Speeding up decision-making: “Looking at datasets, making decisions based on patterns, it helps you go faster.”
• Improving user access reviews: “When users are presented with a ton of data to certify, they often don’t understand it…AI can help translate it into something meaningful.”
• Detecting anomalies: “Good luck asking an analyst to find anomalies in Snowflake AI can do that within hours,” said Andrej.

Pitfalls to Avoid

Both speakers stressed the importance of real-world testing. “Don’t be wowed by a flashy demo,” Andrej warned.

“Production-level performance is like running a marathon – not a 5K.”

Kevin added, “You need to test with your own data. Use read-only access, a time-boxed POC, and contractual guarantees to de-risk.”

Another critical takeaway: AI won’t take your job – but someone who knows how to use it might. Kevin noted, 

“There’s a skills gap in this space. AI is how we augment that, not replace people.”

AI as a Layer, Not a Replacement

For companies hesitant to take on a full IGA overhaul, Andrej offered this advice: 

“Start with the AI layer first. Add visibility and intelligence to your existing setup. Don’t rip and replace – augment.”

Want to see the full conversation and dive deeper? Watch the on-demand recording to learn how AI is reshaping IGA and what steps you can take today to future-proof your identity strategy.

You can also visit our website to learn more about autonomous identity or book a demo!