AIR raised fifty million dollars across two seed rounds — ten million led by Sequoia Capital, then forty million led by Greenoaks Capital, closed within weeks of each other — to build continuous vetting and blocking infrastructure for the tools that enterprise AI agents use. The platform discovers all AI agents running inside an organisation, audits the MCP servers, plugins, and skills those agents are connected to, and blocks tools that fail security checks. The threat model: attackers can poison content that AI agents consume, or compromise tool packages that agents invoke, creating supply-chain attack surfaces that traditional endpoint security does not cover. AIR flags approximately 27 per cent of available agent add-ons and skills as potentially risky. Founders Yair Saban and Niv Hoffman are Unit 8200 veterans with offensive cybersecurity backgrounds; angels include Wiz co-founder Yinon Costica and Cognition president Zach Frankel. The Sequoia plus Greenoaks double bet on a two-seed sequence is unusual and reflects the speed at which enterprise AI agent deployments are creating demand for the governance layer AIR provides.