Exclusive: Vinod Khosla wanted ‘every available dollar’ of Runlayer’s funding round. It just raised $30 million to govern the agent workforce

When longtime tech investor Vinod Khosla heard that Runlayer—the startup trying to become the default infrastructure layer governing how every corporate employee interacts with AI agents—was raising a new round, his response was unambiguous: he wanted “to buy every available dollar of the round.” Now, that conviction is official. Runlayer raised a $30 million Series. A led by Felicis, with Khosla Ventures participating, Fortune learned exclusively. Felicis preempted the round which brought Runlayer’s total capital raised to $42 million. The startup, which launched out of stealth just seven months ago with an $11 million seed from the same two investors, is tackling one of enterprise tech’s most pressing problems. Companies are deploying AI agents everywhere, but no one has a reliable way to see what those agents are doing, control what they access, or stop them from going rogue. Last year, one of Runlayer’s customers discovered that an agent had blown through its entire annual AI-compute budget in a single weekend, running in a loop. Runlayer is essentially a corporate app store and control room rolled into one. Employees can plug in any AI tool—OpenAI’s ChatGPT, Anthropic’s Claude, Salesforce’s Agentforce, custom-built agents—and get a pre-approved way to use them, with company data already connected and guardrails already set. IT and security teams, meanwhile, get a single dashboard showing exactly what every AI agent in the organization is doing, what data it touched, and how much it cost. The platform also surfaces “shadow AI“—tools employees are using without their company’s knowledge—a problem some reports suggest could be as high as 78% of enterprise AI users. “Every employee will delegate their work to swarms of agents. Not as a novelty, and not as a side tool, but as a core part of how work gets done.” CEO Andrew Berman told Fortune. Berman, a three-time founder, previously cofounded Nanit—the