Editorial illustration for AIR Raises USD 50M to Vet AI Agent Skills, Citing Unchecked 'Same Risk
AIR Raises $50M to Vet AI Agent Skills
AIR came out of stealth this week with $50 million in seed funding, split across two rounds that closed within weeks of each other. Sequoia led the first, a $10 million round, and Greenoaks led the second at $40 million, according to CEO Yair Saban. Angel money came in too, from Zach Frankel of Cognition, Wiz co-founder Yinon Costica, Eon co-founder Ofir Erlich, Clay co-founder Varun Anand, Anne Neuberger, Omer Adam, and firms Swish and Netz.
Saban founded AIR with CTO Niv Hoffman. Both are alumni of Unit 8200, the Israeli intelligence corps known for producing offensive cybersecurity talent. Their bet: as companies hand AI agents access to more of their internal systems, a new kind of software supply chain is forming around the skills, plugins, MCP servers, and add-ons those agents pull in from outside. Right now, Saban argues, almost none of that gets vetted before it touches company infrastructure.
AIR's platform is built to find agents running inside a company, check the tools and components they rely on, and cut off anything that doesn't clear its security bar. It also runs a marketplace of pre-vetted add-ons.
Founded by Yair Saban (CEO) and Niv Hoffman (CTO), veterans of Israel’s Unit 8200 intelligence corps, where they worked on offensive cybersecurity, AIR offers a platform that can discover agents running inside companies, continuously vet any skills, tools, and components those agents use, and block them from interacting with software or external sources that don’t pass security criteria.
Why this matters
AIR's $50 million bet is really a bet on how fast we're granting AI agents access without asking who wrote the code they're running. Saban's comparison to open-source supply chain risk is fair, and a little uncomfortable. We spent a decade building tools like Snyk and Dependabot to catch malicious npm packages, and it still took years and several high-profile breaches before enterprises took it seriously. Skills, plugins, and MCP servers are the new dependency tree, except now they're wired into systems that can act on their own, touching databases and live internet connections with far less human review in the loop.
For developers and founders shipping agents right now, the practical takeaway is uncomfortable: most teams have no inventory of which third-party skills their agents call, let alone a way to audit what those skills actually do once installed. AIR is one bet on solving that gap. Whether it or a competitor wins matters less than the fact that this category needs to exist before the first serious incident forces the issue.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Perplexity says it isn't using confidential data for post-training AI
From Archives OpenClaw 2.0 Aims to Make AI Coding a 'Multiplayer' Experience for Teams
Common Questions Answered
What is AIR's core platform designed to do for AI agent security?
AIR's platform discovers AI agents running inside companies and continuously vets any skills, tools, and components those agents use to ensure security compliance. The platform can block agents from interacting with software or external sources that don't meet established security criteria, providing comprehensive oversight of agent capabilities.
Who founded AIR and what is their background?
AIR was founded by Yair Saban (CEO) and Niv Hoffman (CTO), both veterans of Israel's Unit 8200 intelligence corps where they specialized in offensive cybersecurity. Their background in intelligence and cybersecurity informed their approach to addressing unchecked AI agent risks in enterprise environments.
How much funding did AIR raise and which investors led the rounds?
AIR raised $50 million in seed funding across two rounds that closed within weeks of each other, with Sequoia leading the first $10 million round and Greenoaks leading the second $40 million round. The funding also included angel investments from notable figures including Zach Frankel of Cognition, Wiz co-founder Yinon Costica, and several other tech entrepreneurs and firms.
How does AIR's approach to AI agent vetting compare to existing supply chain security tools?
AIR's vetting of AI agent skills, plugins, and MCP servers is comparable to how tools like Snyk and Dependabot address open-source supply chain risks by catching malicious packages. The comparison highlights that enterprises took years and experienced several high-profile breaches before taking dependency security seriously, suggesting similar challenges may emerge with AI agent oversight.
What is the main risk that AIR is addressing in the AI agent market?
AIR is addressing the risk of companies granting AI agents access to tools and data without properly vetting who wrote the code and components those agents are running. This unchecked access represents a significant security vulnerability similar to the supply chain risks that plagued the open-source software ecosystem.
Further Reading
- AIR Security raises $50 million to build a firewall for AI agents - CTech
- AIR raises $50M to help companies vet the skills and add-ons AI agents use - TechCrunch
- AIR raises $50M to build a firewall for AI agents before they go rogue - CryptoBriefing
- Security Evaluation Benchmark for AI Agents - IETF
- A Comprehensive Framework for Evaluating Real-World AI ... - arXiv
