Israeli cybersecurity firm Zenity has raised $125 million in a funding round aimed at expanding its platform for securing AI agents deployed across enterprise environments, according to Globes reporting. The round marks one of the largest capital infusions into the narrowing but fast-growing niche of agentic AI security, a field that has attracted urgent attention as businesses race to deploy autonomous AI systems capable of taking independent actions within corporate networks.

The raise underscores a widening recognition among investors and security professionals that AI agents — software systems that can browse, execute code, query databases, and interact with external services without direct human oversight — introduce a new and poorly understood attack surface. As enterprises integrate these tools into sensitive workflows, the risk of manipulation, data exfiltration, or unauthorized action grows proportionally. That threat landscape is directly relevant to defense-adjacent organizations and government contractors operating under strict data-handling requirements, a concern GDD has tracked in coverage of AI sandbox breaches linked to similar autonomous systems.

What Zenity’s Platform Targets

Zenity focuses specifically on securing what the industry calls agentic AI — systems built on large language models that are granted permissions to act on behalf of users or organizations. These agents can be embedded in productivity platforms, customer-service pipelines, or internal enterprise tools, and they typically operate with elevated access to sensitive data and third-party integrations. The attack vectors Zenity addresses include prompt injection, where a malicious instruction is embedded in content the agent processes, and privilege escalation, where an agent is manipulated into performing actions beyond its intended scope.

The company’s approach centers on monitoring agent behavior in real time, enforcing policy guardrails, and providing visibility into what autonomous systems are actually doing once deployed. That kind of runtime observability is largely absent from standard enterprise security stacks, which were not designed with agentic workflows in mind. Zenity’s pitch is that as AI agents proliferate, the gap between what an organization believes its AI is doing and what it is actually doing will become a primary vector for breaches.

Investment Context and Market Trajectory

The $125 million raise positions Zenity among the better-capitalized pure-play AI security startups, at a moment when both the commercial and national-security communities are grappling with how to govern AI systems that increasingly operate beyond direct human supervision. The Globes report does not identify the lead investor or break down the round by tranche, and the company’s current valuation was not disclosed in the reporting.

The broader funding environment for AI security remains competitive. Enterprises are under pressure from regulators and insurers alike to demonstrate that their AI deployments meet basic security standards, and the absence of mature tooling in the agentic space has created commercial opportunity. For vendors serving government or defense-sector clients — where data compartmentalization and auditability are non-negotiable — the ability to enforce behavioral policies on AI agents in real time is not a luxury feature but an operational requirement. Zenity’s funding suggests investors believe that requirement will translate into durable enterprise demand well beyond the current AI deployment wave.