AI agents are moving from experiments to real business systems — but how do enterprises actually govern machines that can act autonomously?

In this episode of the Chief AI Officer Podcast, Sanjay Puri speaks with Blake Brannon, Chief Innovation Officer at OneTrust, about one of the hardest problems in enterprise AI: how to build trust and control when AI agents can take actions across business systems, data and workflows.

Blake draws a powerful parallel between today’s AI transformation and the mobile and cloud era — arguing that enterprises need to shift their focus away from simply trusting devices, models or AI harnesses and toward governing what an AI agent actually does.

The conversation explores why AI agents inherit a user’s identity but not their trust, why traditional AI guardrails may not be enough, and why the most important control point may be the action itself — the tool call connecting an agent to enterprise systems.

They also discuss the growing accountability burden on enterprises, the gap between AI governance policies and real-time enforcement, shadow AI hidden inside third-party software, AI risk and insurance, ISO 42001, and the talent shortage facing organizations trying to build effective AI governance.

In this episode:

  • Why AI governance needs to learn from the mobile and cloud era
  • Why enterprises should govern AI agent actions, not just AI models
  • The “guardian agent” approach to governing autonomous systems
  • Why AI guardrails alone cannot be trusted
  • How AI agents inherit identity but not human accountability
  • The shift in liability from AI model providers to enterprises
  • Why having an AI governance policy is not the same as enforcing it
  • The gap between policy intent and runtime controls
  • What happens when AI agents outnumber humans in the enterprise
  • The rise of shadow AI inside third-party software
  • How enterprises can manage AI features introduced through software updates
  • Can AI governance become an enabler of faster AI adoption?
  • AI risk, cyber insurance and the emerging market for AI assurance
  • Why intent and purpose matter when governing autonomous AI actions
  • The talent challenge at the intersection of AI, privacy, cybersecurity and compliance

The central question is no longer simply: “Can we trust this AI model?”

It is: “Can we trust what this AI agent is about to do?”

For enterprises deploying AI at scale, that distinction could define the next generation of AI governance.

Chief AI Officer is a candid conversation series bringing you inside the minds of the leaders responsible for turning enterprise AI from strategy into reality.

Subscribe for conversations with Chief AI Officers, technology leaders and the people shaping the future of enterprise AI.

#ChiefAIOfficer #AI #ArtificialIntelligence #AIAgents #AIGovernance #EnterpriseAI #AILeadership #ResponsibleAI #Cybersecurity #AITrust #OneTrust #AITransformation