A company that just secures AI agents raised $125m — the trust gap in agentic coding is now a venture-scale market
AI agent security and governance startup Zenity raised a $125m Series C on 3 August 2026, led by Norwest with SoftBank Vision Fund 2 and Intel Capital among the backers, after tripling revenue for two straight years — evidence that the risk of letting autonomous AI agents take real actions in production is now a funded category, not an edge case.
4 September 2026
Zenity, a platform built to govern what autonomous AI agents are allowed to do, closed a $125m Series C on 3 August 2026, led by Norwest with new investors including SoftBank Vision Fund 2, Qumra Capital, Hitachi Ventures and LG Technology Ventures joining existing backers Vertex Ventures and Intel Capital. The company says it has tripled revenue in each of the past two years and serves Fortune 500 and Global 2000 customers in financial services, healthcare and pharma. Its pitch isn’t model safety or prompt filtering — it’s runtime control: deciding whether to allow, modify or block a specific agent action, across Microsoft Copilot, ChatGPT, Gemini, Claude and custom agents running in AWS and Google Cloud, before that action executes.
Why this lands differently to a typical AI funding story
We’ve tracked the “trust gap” in agentic coding for months — surveys showing developers don’t fully trust AI-generated code, security researchers finding sandbox escapes and prompt-injection routes in coding agents, and a steady run of incidents where an agent was talked into doing something it shouldn’t. Zenity’s raise is the commercial answer to that gap: enterprises are now paying for infrastructure whose whole job is assuming an AI agent will eventually be tricked or will misjudge an action, and stopping the blast radius before it matters. A $125m round, on tripled revenue, for a category that didn’t meaningfully exist three years ago, is a clear signal that “let the agent just do it” is no longer an acceptable default anywhere agents touch production systems, data or money.
So what
If your team — or a vendor building for you — is letting AI coding agents commit, deploy, or act on production systems with real autonomy, the market is telling you that unmanaged agent permissions are now treated as an operational risk worth funding a whole company to solve. Good AI-assisted development means agents that accelerate delivery inside a workflow a human actually controls and reviews, not agents given the keys. See how we build with AI-assisted development practices that keep that control in place, or get in touch to talk through how it applies to your build.