_(2).jpg)
The adoption of AI agents built with platforms like OpenAI's AgentKit is accelerating, but this innovation introduces significant new security challenges. Zenity, a leader in AI agent security, is addressing these concerns head-on with a new solution designed to provide deterministic protection where native guardrails may fall short.
Zenity has launched runtime protection specifically for AI agents built with OpenAI's AgentKit.
The solution provides endpoint-level inspection of every user-agent interaction in real time.
It closes critical security gaps identified by Zenity Labs, including prompt injection and data leakage.
Core capabilities include blocking data exfiltration, preventing secrets exposure, and stopping unsafe responses.
Unlike probabilistic guardrails, Zenity enforces deterministic, rule-based security policies.
This gives enterprises a way to securely scale their use of AgentKit for internal and customer-facing workflows.
This launch is a direct response to research from Zenity Labs that exposed critical vulnerabilities in OpenAI's AgentKit guardrails. These security gaps can be bypassed through sophisticated attack vectors like prompt injection, response obfuscation, and credential exposure. As enterprises begin deploying AgentKit for critical workflows, the potential for data leakage and non-compliant agent behavior becomes a top concern for security teams.
Zenity's runtime protection operates at the endpoint level, inspecting every interaction between users and agents built with AgentKit. The key differentiator is its shift from probabilistic to deterministic security. Instead of relying on AI-based judgments that can be misled, Zenity applies clear, predefined security rules to ensure predictable and enforceable protection. This method allows it to identify and block risky behavior before a response ever reaches the end-user, effectively closing the gaps left by native guardrails.
The platform delivers several critical security functions for organizations using AgentKit. Its Data Leakage Detection identifies and blocks attempts by agents to exfiltrate sensitive or regulated information. Furthermore, it provides Secrets Exposure Prevention by detecting embedded credentials or API keys in agent responses and blocking those actions. Finally, its Unsafe Response Blocking prevents interactions that violate corporate policy, compliance standards, or brand trust from being delivered to the user.
“AgentKit accelerates how AI agents are built and scaled, but it also expands the attack surface overnight,” said Michael Bargury, CTO and co-founder of Zenity. “Our research shows that AgentKit’s guardrails can miss critical risks - from subtle prompt injections to hidden data leakage. Zenity’s runtime protection closes that gap by inspecting every response, understanding intent and enforcing security policies.”
With this launch, Zenity provides security teams with a crucial tool for managing the risks associated with agentic AI, enabling safer and more compliant innovation as adoption of platforms like OpenAI's AgentKit continues to grow.
Zenity is the first security and governance platform purpose-built for AI agents - spanning SaaS, home grown platforms (Cloud), and end-user devices (Endpoint). Trusted by Fortune 500 enterprises, Zenity helps security teams confidently adopt AI by delivering defense in depth with full-lifecycle coverage: from agent discovery and posture management to real-time detection, inline prevention, and response. With an agent-centric approach that prioritizes how agents behave, what they access, and which tools they invoke, Zenity eliminates blind spots and enforces consistent policy and controls across environments so organizations can innovate with AI, without compromising security.
TechIntelPro empowers B2B technology decision-makers with the latest industry trends, deep insights, leadership perspectives and actionable intelligence to drive both organizational and professional milestones—and thrive on the tech pulse.