Securing Enterprise AI in Practice: Why Governance Starts and Ends with Agents

AI agent security incidents are no longer anecdotal vulnerabilities - they present a persistent risk surface that threatens traditional identity, application, and network security as agents span across workflows. At the same time, agents have proven to be where organizations can drive the most value from their AI development.

According to CSA’s April 2026 AI Agent Security Survey, 74% of enterprises expect their organizations will have over 100 agents live by the end of 2026. Meanwhile 53% of participants noted that agents exceeded intended permissions or acted out of scope, and 47% experienced a security incident involving an agent in the last year.

On our June 15 webinar, join Zenity experts Chris Hughes, Rock Lambros, and Ben Hanson to learn:

• Why agents are where enterprise AI gets put into action, and how it becomes consequential when security isn’t purposefully considered

• Where siloed model, identity, endpoint, and application security practices fall short when preventing or remediating the ephemeral and changing attack surfaces agents pose

• Why frameworks like the OWASP Top 10 for Agentic Applications have become relevant, and how to put them into practice

• How “secure-by-design” comes up short in practice, and how to make it a reality

Chris Hughes

VP of Security Strategy

Ben Hanson

Field CTO & Director of Field Engineering

Rock Lambros