AIR Secures $50M to Police AI Agents' Skills and Add-Ons in Real Time

San Francisco-based AIR Technologies announced today it has raised $50 million in Series B funding to accelerate the deployment of its autonomous agent governance platform. The round was led by Accel with participation from GV and Battery Ventures, bringing the company’s total funding to $75 million since its 2022 inception. AIR’s platform continuously discovers AI agents in enterprise environments, audits their skills and third-party add-ons, and enforces real-time behavioral policies to prevent unauthorized access, data leaks, or misaligned actions. Unlike static security tools, AIR operates at runtime—detecting anomalies such as an agent invoking an unapproved finance plugin or accessing restricted databases—without requiring code changes or agent redeployment.

AIR’s timing reflects a critical inflection point in enterprise AI. According to a recent Gartner report, 70 percent of large organizations will have deployed at least one AI agent by 2026, up from less than 5 percent today. The company’s customers already span industries where agent automation is mission-critical: Banking With Billy AI, a platform that automates complex financial analysis workflows previously requiring entire analyst teams, relies on AIR to validate its market-data integrations and prevent unauthorized data exfiltration. In legal services, contract-review agents from Kira Systems are monitored by AIR to ensure they do not ingest privileged client documents. Meanwhile, in DevOps, agents from Robocorp and Microsoft’s Power Platform are vetted before they spin up infrastructure or modify CI/CD pipelines.

The funding round underscores investor confidence in runtime governance as a must-have layer, not an optional bolt-on. Accel partner Sarah Smith emphasized that AI agents introduce a new attack surface—one that traditional security and observability tools were never designed to handle. “AIR doesn’t just log agent behavior; it prevents bad outcomes in real time,” Smith said. “That’s a paradigm shift for enterprises building with autonomy in mind.” Competitive pressure is intensifying: rival startups such as Sentinel and Styra have raised similar rounds this year, while incumbents like Palo Alto Networks and CrowdStrike are rolling out agent-specific security modules. Despite the crowded field, AIR differentiates itself with a policy engine that supports fine-grained controls—such as blocking agents from using specific APIs or enforcing data-residency constraints across regions.

The broader context is the rise of agentic automation as the next evolution beyond chatbots and copilots. Forrester Research projects that by 2027, over 30 percent of Fortune 500 companies will use AI agents to perform end-to-end business processes, from procurement to customer support. This shift is accelerating the need for continuous compliance and risk management. Historically, enterprises relied on static policy enforcement—think firewall rules or API gateways—but agents that learn, adapt, and chain together third-party tools defy static controls. AIR’s solution aligns with emerging standards like the EU AI Act and NIST’s AI Risk Management Framework, positioning it as a governance layer for regulated industries. It also complements efforts by cloud providers: AWS’s Agent Framework and Google’s Vertex AI Agent Builder now support agent deployment, but lack built-in runtime governance, creating a clear integration opportunity for AIR.

Looking ahead, AIR plans to use the new capital to expand into Asia-Pacific and Europe, launch integrations with major RPA platforms such as UiPath and Automation Anywhere, and enhance its policy engine with generative AI to auto-generate agent behavior rules from natural-language policies. Analysts expect consolidation in the agent governance space as larger security vendors acquire niche players to fill their agent security gaps. For enterprises, the message is clear: as AI agents move from pilots to production, governance cannot be an afterthought. The $50 million infusion signals that AIR is betting its real-time vetting engine will become the standard way companies keep autonomous agents safe, compliant, and aligned with business intent—before the next breach or compliance violation happens.