$125M Bet on AI Agent Security: A New Frontier Emerges

Israeli AI security startup Zenity has secured a substantial $125 million in a Series C funding round. The investment was led by Norwest Venture Partners, with participation from SoftBank Vision Fund 2, Hitachi Ventures, and LG Technology Ventures. This major capital infusion highlights a growing investor focus on the novel security challenges arising from the enterprise adoption of autonomous AI agents.

The Shift: From Model Security to Agent Behavior Security

Historically, AI security efforts concentrated on model vulnerabilities, adversarial attacks, and screening user prompts. However, as AI agents begin to perform real tasks—making autonomous decisions, taking actions, and connecting directly to core business systems—the nature of the threat landscape is evolving.

Ben Kliger, Co-founder and CEO of Zenity, explains that the core risk is shifting from the “model and its input” to the “agent’s behavior itself.” Even a properly trained model can, through its autonomous operations, exhibit behaviors that diverge from corporate intentions, potentially leading to data exposure, asset loss, or system disruption.

Real-Time Monitoring and Intervention: The Zenity Approach

At its core, the Zenity platform acts as a “safety rail” for enterprise AI agents. It does not replace existing AI models but instead continuously monitors an agent’s operational steps within a live environment to analyze its behavioral trajectory.

  • Behavioral Baseline Comparison: The system establishes a normal behavior baseline for each agent, constantly comparing current actions against intended goals.
  • Real-Time Risk Assessment: Upon detecting anomalous behavior that could threaten corporate data, customer assets, or critical systems, the platform performs an immediate risk evaluation.
  • Dynamic Blocking and Adjustment: Depending on the risk level, the platform can automatically halt dangerous operations or dynamically adjust the agent’s parameters to steer it back to a safe course of action.

This approach functions like a 24/7 “security supervisor” for autonomous AI workers, ensuring their actions remain aligned with enterprise security policies even as they operate independently.

Capital Backs the Future: An Inevitable Enterprise Need

This significant funding round validates not only Zenity’s technology but also the market’s anticipation of the AI Agent security sector’s growth potential. As businesses across industries accelerate the AI-driven automation of their processes, autonomous agents will become ubiquitous. Ensuring these “digital employees” operate safely, reliably, and in compliance will form a foundational layer of enterprise digital infrastructure, representing a rapidly emerging market.