Runlayer raises $30M to govern AI agents
Runlayer raised a $30M Series A led by Felicis with Khosla Ventures to build a security and control layer for the MCP agents companies are deploying.
Runlayer, a New York startup building security infrastructure for AI agents, raised a $30 million Series A on June 24. Felicis led the round, with Khosla Ventures participating, taking the company’s total funding to $42 million. It launched out of stealth about six months ago.
The product sits around the Model Context Protocol, the open standard for how agents connect to company systems and data. Runlayer gives security teams identity management, scoped permissions, policy enforcement, audit logs, and real-time visibility into what each agent touches and how much it costs. The pitch is a single control plane so IT can approve agents and tools rather than discover them after the fact. Named customers include Instacart, Gusto, Decagon, Opendoor, dbt Labs, AngelList, and Lemonade.
What it means for you
The money is following a real gap. Teams are wiring agents into email, code, and internal tools faster than anyone can track what those agents are allowed to do. A governance layer is how a small operator says yes to AI without handing it the keys to everything.
This is the same thread running through enterprise MCP managed auth and the funding for agent identity startup NewCore. If you are wiring up agents yourself, the practical starting point is giving your AI agents their own keys.