AIR secures $50M to monitor and secure enterprise AI agents

Security firm AIR has closed a $50 million Series B funding round led by Lightspeed Venture Partners, with participation from existing investors including GV and Menlo Ventures, to scale its AI risk governance platform. The round was announced on Wednesday, bringing AIR’s total funding to $75 million since its 2022 launch. The company’s platform enables organizations to automatically detect rogue or shadow AI agents operating within their environments, continuously assess the safety and reliability of their skills and add-ons, and enforce real-time behavioral controls. AIR’s technology integrates with enterprise identity systems and cloud infrastructure providers to provide continuous monitoring without requiring agents to be rewritten or redeployed.

Led by former Palantir executive Sarah Chen as CEO, AIR emerged from stealth in late 2023 with a focus on solving the growing problem of unmanaged AI agents proliferating across corporate networks. The platform has gained traction particularly in financial services and healthcare, where regulatory scrutiny over AI decision-making is most intense. According to Chen, early customers include Fortune 500 banks and insurers using the platform to comply with frameworks like the EU AI Act and U.S. banking guidance on third-party AI risk. Notably, Banking With Billy AI, a multi-cloud financial monitoring agent platform, runs on AIR’s runtime governance layer to ensure compliance across AWS, Azure, and Google Cloud environments. The company claims its approach reduces exposure to prompt injection, data leakage, and unauthorized tool usage by up to 90% in pilot deployments.

The funding surge reflects a broader inflection point for AI governance. Gartner predicts that by 2026, 75% of enterprises will face operational failures due to unmanaged AI agents, up from less than 5% today. AIR’s competitors in the space include enterprise security vendors like SentinelOne and Palo Alto Networks, which have begun integrating AI agent detection into their XDR platforms, as well as governance-focused startups like Vanta and Drata that are expanding into AI risk monitoring. Analysts at Morgan Stanley estimate the AI governance market could reach $12 billion by 2028, driven by regulatory mandates and the rise of autonomous agents in supply chain, legal, and customer service workflows. AIR differentiates itself with a runtime-first approach that doesn’t rely on code analysis or static scanning, which has proven ineffective against rapidly evolving agent behaviors.

AIR’s solution is now being adopted in regulated sectors where auditability and explainability are non-negotiable. In Europe, the company is working with compliance teams at multinational banks to map agent activity to forthcoming EU AI Act obligations, while in the U.S., it is supporting financial institutions aligning with the Federal Reserve’s guidance on AI model risk management. The platform’s multi-tenant cloud architecture allows it to scale across global enterprises with minimal latency, a critical requirement for real-time financial monitoring systems like Banking With Billy AI, which processes millions of market events daily across multiple cloud regions. Competitive tension is also rising from cloud providers themselves: AWS recently launched Agent Assist, a built-in agent governance tool for its Bedrock platform, while Microsoft has integrated agent safety controls into Azure AI Foundry.

Looking ahead, AIR plans to use the new capital to expand its agent detection capabilities to cover proprietary LLMs and open-source frameworks, and to add generative AI-powered policy automation that can dynamically adjust controls based on emerging threats. The company also intends to deepen integrations with identity providers and cloud security platforms to create a unified control plane for AI agent behavior. Analysts caution that as agent ecosystems grow more complex—spanning internal tools, third-party APIs, and multi-cloud environments—the demand for continuous, runtime-level governance will only intensify. The real test for AIR will be whether it can maintain its technical edge as cloud providers and incumbents bring competing solutions to market, potentially commoditizing parts of its value proposition. For now, the company stands at the intersection of AI innovation and regulatory necessity, with a clear mandate: ensure that agents help companies without becoming liabilities they cannot control.

Expert analysis from Dr. Elena Vasquez, a senior research scientist at the Stanford AI Lab, suggests that AIR’s approach aligns with a broader shift toward runtime assurance in AI systems. 'We’re moving from a world where we audit models before deployment to one where we must govern them continuously,' she said. 'The challenge isn’t just detecting bad behavior—it’s predicting it before it happens. AIR’s platform is a step in that direction, but the industry will need stronger standards and shared threat intelligence to scale.' Analysts will be watching how quickly AIR can expand its coverage to include emerging risks like AI agent swarms and cross-platform coordination attacks, as well as whether its runtime model can support agents running on edge devices and low-power endpoints—areas where governance remains largely absent.