MCP AI agent security startup Runlayer launches with 8 unicorns, $11M from Khosla’s Keith Rabois and Felicis
What Happened
Three-time founder Andrew Berman is back with a startup that helps IT ensure business users' AI agents operate securely.
Our Take
Agent security is real. As orgs deploy AI agents with production access, the blast radius explodes—prompt injection, hallucinated API calls, uncontrolled decisions. Runlayer's bet on MCP + Claude is smart (tight niche, aligned with Anthropic's roadmap). $11M from Khosla and Rabois validates something.
But here's the gap: the product's still vapor. 'Help IT ensure agents operate securely' is too vague. Monitoring? Sandboxing? Access controls? Until they ship and prove market fit, they're a thesis, not a company.
What To Do
Watch if you're deploying Claude agents at scale, but wait for a product demo before committing.
Cited By
React