Okta Streamcast Episode 2: The shadow AI takeover

Your users are deploying AI agents to move faster. But every autonomous agent introduces a new, often invisible, attack vector. Many run with long-lived credentials, operate outside your IT controls, and access sensitive data without clear ownership.

In this short, no-slides conversation, Okta and Zscaler explore how Shadow AI becomes a real breach risk—and how to shut that door.

What you’ll walk away with:

• A proven approach to discovering and mapping every AI agent in your environment and expose hidden actors
• How to establish a single source of truth for autonomous agent identities.
• Practical guidance on governing agent access across its full lifecycle, from creation to revocation
• Insights from Okta and Zscaler on blocking risky agent behavior in real time
• Actionable ideas to eliminate the AI governance gap and secure autonomy at scale
  

Speakers

Harish Peri, SVP & GM, AI Security, Okta
Michael Schofield, Director, Field Product Management - AI Security, Zscaler