AIR, an AI agent security startup founded by Yair Saban and Niv Hoffman, came out of stealth with $50 million raised across two seed rounds. Sequoia Capital led the first $10 million round and Greenoaks Capital led a second $40 million round weeks later, according to TechCrunch and Calcalist.
The platform discovers AI agents running inside a company, continuously vets the skills, plug-ins and MCP servers those agents use, and blocks ones that fail its security checks before they reach an agent's decision making context, both outlets reported. AIR said it filters out roughly 27% of the add-ons and skills it evaluates.
"In the early 2000s, whenever you installed a driver, the driver didn't need to be signed... You don't have that with skills or plug-ins or MCPs," CEO Saban said, according to TechCrunch. Sequoia partner Bogomil Balkansky called it "not a scanning problem, it is a continuous re-verification problem," TechCrunch reported.
AIR's own research found more than 17,800 public AI add-ons, with about 6.7 million installations, that rely on untrusted external sources for their instructions, Calcalist reported. The company has 40 employees and more than 20 customers, about a quarter of them large enterprises concentrated in financial services and pharmaceuticals, according to TechCrunch.
This is the supply chain problem the software industry already solved once, for package managers and browser extensions, showing up again for agent skills and MCP servers. Any team wiring third party skills or MCP servers into an agent without a vetting layer is running the same unsigned-driver risk AIR is describing.