About 98% of organizations surveyed experienced at least one AI agent-related incident that disrupted business or the organization.

So what happens as companies hand AI agents more autonomy, access, and decision-making power?

In this episode of Work 2.0: The Podcast, Rachel Wells sits down with Kavitha Mariappan, Chief Transformation Officer at Rubrik, and Vaibhav Sahgal of Economist Enterprise to examine new research into the risks emerging as agentic AI moves deeper into organizations.

They unpack what an AI agent incident actually looks like, from unauthorized access and sensitive data exposure to corrupted business data, cascading agent-to-agent actions and threats that can spread at machine speed.

The conversation goes deeper into a problem many organizations are only beginning to confront: What happens when an autonomous AI system has already acted before a human can intervene?

Rachel, Kavitha, and Vaibhav explore:

* Why 98% of surveyed organizations reported disruptive AI agent-related incidents

* What happens when AI agents operate beyond their authorized permissions

* Why traditional enterprise security controls may struggle with autonomous AI

* Whether an AI “kill switch” can actually stop the damage

* The difference between stopping an AI agent and reversing what it has already done

* Why companies need visibility into every AI agent operating across the organization

* How least-privilege access, data minimization, and anomaly detection can reduce the blast radius

* Why third-party AI systems could become a major enterprise vulnerability

* What “assume breach” means in an AI-powered organization

* Whether companies could eventually be required to disclose major AI agent incidents to regulators, investors and customers

We discuss how organizations can remain resilient when software can reason, act and interact with other systems at machine speed.

Work 2.0: The Podcast examines AI, power and the changing economics of work, bringing researchers, executives and technology leaders into conversations about what technological change actually means for organizations and the people inside them.

Follow the show for more conversations on artificial intelligence, agentic AI, the future of work, leadership, organizational transformation and the changing relationship between humans and intelligent systems.

Featured guests: Kavitha Mariappan, Chief Transformation Officer, Rubrik; Vaibhav Sahgal, Economist Enterprise

Economist Enterprise produced the research discussed in this episode in collaboration with Rubrik.

🎧 Available now on Spotify, Apple Podcasts, YouTube Podcasts, and more.

Don’t forget to rate, review, and share this episode with a co-worker or friend.

Get full access to Work 2.0 at therealrachelwells.substack.com/subscribe