When we started Onyx, the question I could not stop thinking about was this: when something smarter than you is responsible for the things you depend on, how do you stay in control?
Two years later, that question is no longer hypothetical.
A year ago, almost nothing inside the enterprise was done by an agent. A year from now, most actions inside the enterprise will be. And not long after that, entire companies will run as autonomous agent operations, making the decisions humans used to make about capital allocation, energy grids, patient care, and supply chains.
The models underneath those agents are crossing into the realm of general intelligence. Their capability at this level makes AI a national security matter, and it is arriving faster than the controls around it. The gap between what advanced AI can do and what humans can govern is widening, not narrowing. The upside is real: abundance, resilient supply of food and energy. The risk is that we delegate those decisions to systems no one is watching, on infrastructure no one is governing.
Humans need to stay in control of advanced AI. Not by holding it back. By governing it well enough to safely let it further into the fabric of how we run critical industries.
That is why we built the Secure AI Control Plane. And that is why, today, I am announcing that Onyx has raised $113 million in Series B funding led by Bessemer Venture Partners, with participation from Cyberstarts, TCV, Conviction, FirstMark, Vintage Investment Partners, QuantumLight, and G Squared.
What the round is for
The Series B is not about market share inside a category that already exists. It is about building the platform that lets humans stay in control as agents take on more of the work.
Every enterprise is becoming an agent operator, whether they planned to or not. Agents now operate across the browser, endpoint, SaaS, and cloud surfaces where enterprises are standing up their AI programs. The Secure AI Control Plane governs the agent where the actor lives. It discovers the AI agents in the environment and inspects every action before it takes effect, without slowing adoption down.
Onyx already secures more than 1.1 million agents across our enterprise customers and inspects more than 66 million AI sessions in real time. That footprint is what makes the control problem tractable, not hypothetical.
Where this work is most urgent
The industries that carry the weight of critical infrastructure need this first. Energy, financial services, healthcare. These are the sectors where an agent making the wrong decision at production speed does not produce a headline; it produces an outage, a market event, or a life at risk. The Fortune 500 companies we work with across these industries are the ones setting the pace on how agent governance actually gets deployed. The depth we owe them is the reason for the Series B.
I want to thank our customers for trusting Onyx with the AI programs, our investors, and the entire team at Onyx who helped turn conviction into a control plane a CISO can actually deploy inside their business, on the timeline the business needs.
The question I started with is still the question I am working on. When something smarter than you is responsible for the things you depend on, how do you stay in control?
We are answering it. And we are moving quickly, because advanced AI is not slowing down for the security controls around it to catch up.
Frequently Asked Questions
The Secure AI Control Plane is Onyx Security's platform for governing enterprise AI agents across browser, endpoint, SaaS, and cloud surfaces. It discovers AI agents in an environment and inspects every action before it takes effect – without slowing adoption. Onyx currently secures more than 1.1 million agents and inspects over 66 million AI sessions in real time.
Traditional AI security focuses on protecting models and data at rest. Agent governance operates at runtime, inspecting every prompt, response, and agent action as it happens across production environments. This runtime approach addresses the control gap created when autonomous agents make decisions at production speed in critical industries like energy, healthcare, and financial services.
Onyx governs AI agents where the actor lives – across browser, endpoint, SaaS, and cloud surfaces. The platform discovers agents operating in the environment and inspects each action before it takes effect. This runtime inspection covers every prompt, response, and agent action, enforcing controls without introducing friction that slows enterprise AI adoption.
Without agent governance, autonomous AI agents making decisions at production speed in critical infrastructure sectors can cause outages, market events, or endanger lives. The Onyx Security Series B funding of $113 million reflects growing enterprise demand for controls that let organizations safely scale AI adoption while maintaining human oversight across capital allocation, energy grids, patient care, and supply chains.
Enterprises should prioritize platforms that discover agents across every surface – browser, endpoint, SaaS, and cloud – and enforce controls at runtime without blocking adoption. Critical infrastructure industries like energy, financial services, and healthcare need this capability first. Fortune 500 companies in these sectors are setting the pace for how agent governance gets deployed.
