Onyx Security launched Thursday with $40 million in funding from Conviction Partners and Cyberstarts to build what it calls a secure AI control plane to expand the use of autonomous AI agents within enterprises, according to a company announcement.
As companies increasingly deploy AI-driven security and governance systems across functions such as engineering and customer service, the company said its platform is designed to discover AI agents, monitor their reasoning steps, and approve, block, or remediate actions in real-time.
Onyx said the rise of AI agents is creating new operational and security risks for enterprises, including attacks via reasoning errors, hallucinations, and malicious prompts. The company argues that traditional security tools are built for predictable software and human-driven workflows rather than autonomously operating systems, and that this shift reflects broader industry concerns about AI-native cyber defenses.
Maxime Barr Kogan, CEO and co-founder, said companies are already becoming “agent-operators”, giving AI systems access to sensitive internal platforms and critical business processes. The company said its products are intended to provide security, governance, and infrastructure teams with a shared layer of monitoring and policy enforcement.
Onyx was founded by former Unit 8200 member Bar Kogan and veteran AI researcher Gil Elbaz, who previously reported to Nvidia’s CTO and also worked at the IDF AI Research Unit. The company, which already works with multiple Fortune 500 customers and employs more than 70 people in Israel, the United States and Canada, said it plans to use the funding to expand its product and engineering teams.
Cyberstarts Partner Hila Zigman said AI systems are becoming part of organizations’ operational infrastructure, creating what she describes as “a whole new attack surface.” This framework reflects the broader industry view that the adoption of agents is moving faster than the controls to manage them.
