The new company Onyx Security is building an AI control plane that monitors, controls, and corrects autonomously operating AI agents in business environments.
Onyx Security officially launched today, raising a $40 million funding round from investors Conviction and Cyberstarts. The company is addressing an increasingly urgent security issue: autonomous AI agents operating without sufficient control within corporate environments.
Behind Onyx are co-founders Maxim Bar Kogan and Gil Elbaz. Bar Kogan has a background with the Israeli intelligence unit Unit 8200. Elbaz previously worked as an AI researcher at Nvidia and in the Israeli army’s AI unit.
“Every enterprise is becoming an agent operator — whether they planned to or not,” Bar Kogan said at the launch. Agents have access to the most critical systems, but there is no guarantee they will not make serious mistakes or be compromised.
One platform for security, governance, and infrastructure
The Onyx platform works with its own supervisory agents and AI models that understand AI reasoning and can intervene in real time. It can block, correct, or forward actions to a human for approval. The platform continuously discovers active AI agents and tracks every step in their reasoning.
This gives security teams visibility, posture management, and runtime protection. Governance teams can demonstrate compliance with security standards and regulatory requirements. Infrastructure teams support AI agents and MCPs while optimizing for cost and latency. Executives can use Onyx to measure AI adoption and track goals.
Tip: Founders Palo Alto and SentinelOne launch security company Cylake