Zenity Launches Runtime Protection for AI Agents Built with OpenAI AgentKit

Why It Matters

By securing the execution of autonomous AI agents, Zenity addresses growing enterprise concerns over rogue behavior and data leakage, potentially setting a new baseline for AI‑agent safety across regulated industries.

Summary

Zenity announced a new Runtime Protection solution designed specifically for AI agents built on OpenAI’s AgentKit. The offering adds a lightweight security layer that monitors agent behavior in real time, enforces policy controls, and blocks malicious or out‑of‑bounds actions without degrading performance. Zenity says the product integrates natively with AgentKit’s SDK, enabling developers to add protection with a single API call and providing visibility through a centralized dashboard. Early adopters in fintech and healthcare are piloting the service to meet heightened regulatory and risk‑management requirements.