AIR Security launches AI agent firewall with $50M funding

AIR Security has launched from stealth operations with $50 million in funding and a new firewall built specifically for AI agents. The funding round was led by Sequoia Capital and Greenoaks, with participation from several prominent individual industry angels.

According to the company, its research identified more than 17,800 public AI add-ons, representing 6.7 million installations, that rely on untrusted external instruction sources. The firm also discovered AI Skills in the wild that impersonate companies like Anthropic and OpenAI, designed to bypass security reviews and execute arbitrary code.

The need for an AI agent firewall

AI agents are increasingly autonomous, connecting to various tools, data, and third-party services. They browse websites, access files and emails, and act on behalf of employees. This growing autonomy creates a security problem when adversaries influence them through poisoned content or direct compromise, potentially leading to data theft, fraud, or unauthorized access while offering security teams little visibility.

Yair Saban, co-founder and CEO of AIR, stated the core problem. "Every enterprise has a firewall protecting its network. Now they need one protecting their AI agents," he said. He argues that AI agents require a new kind of firewall that protects what enters their context, as they autonomously install tools and connect to internal systems without oversight.

How the AIR firewall operates

The AIR firewall discovers and evaluates every skill, plugin, MCP server, and add-on across an organization's AI agent supply chain, both before and after deployment. Before any third-party or internal add-on is allowed to interact with an enterprise agent, the system performs a deep analysis across known agentic attack patterns. It screens for external instruction sources, hidden behaviors, and typo-squatted packages masquerading as official developer tools.

If an add-on is found to be malicious, vulnerable, or unapproved, security teams can trace every agent and workflow that depends on it and revoke its access across the organization. This process is continuous. If a maintainer pushes a malicious update or an existing integration is compromised later, trust is automatically revoked.

Founding team and enterprise concerns

Saban partnered with Niv Hoffman, who serves as CTO, to found AIR. They were joined by Ryan Knisley, former CISO at The Walt Disney Company and Costco Wholesale, as chief strategy officer. The company positions its solution as a necessary "seatbelt" for enterprises rapidly adopting coding agents like Claude Code, Cursor, and Codex but are afraid to deploy them without security controls.

Through its continuous evaluation of agentic activity across its customer base, AIR also offers a marketplace of pre-vetted, certified add-ons. This provides a safe route for customers to expand their AI agents' capabilities without introducing unmanaged risk.

The company describes AI agents as the new operating system, with AI add-ons as the new applications, marking a fundamental shift in how enterprises will operate and make decisions.