Ben Kliger: Securing enterprises when software starts acting on its own
Summary
- Zenity co-founders Ben Kliger & Michael Bargury shifted focus from no-code security to protecting AI agents, which can access data, make decisions, & act autonomously.
- AI adoption occurred rapidly, outpacing the 10-year growth of public cloud, requiring a redefinition of the traditional cybersecurity perimeter.
- AI agents operate outside single locations, interacting with external systems & removing the concept of a fixed network boundary.
- Concrete risks include agents being manipulated to steal sensitive data, deleting production data, or attacking other organizations.
- Zenity advocates for "human on the loop" strategies to provide intervention points without hindering AI autonomy.
- Kliger urges companies to maintain full speed in AI adoption while implementing dynamic guardrails to fight AI threats with AI.
Zenity CEO Ben Kliger explains why agentic AI security needs a new playbook and how AI guardrails let enterprises deploy agents without losing control.
- Keywords
- AI Stack