Israeli Cybersecurity Firm Onyx Security Raises $113 Million to Secure AI Agents' Actions

Israeli cybersecurity startup Onyx Security announced a $113 million Series B funding round led by Bessemer Venture Partners, just four months after raising $40 million. The company’s valuation now exceeds $640 million. Founded in 2024 by Maxim Bar Kogan, a former Unit 8200 officer and Mixtiles executive, and AI entrepreneur Gil Elbaz, Onyx develops a platform that enables organizations to securely manage AI agents by controlling permissions, monitoring actions, and detecting risks from autonomous operations.

Onyx’s technology tracks the reasoning processes of AI agents to identify and correct malicious or unintended behaviors. The newly raised funds will be used to train the next generation of AI models and expand business development and marketing efforts in the US and other markets. The company employs over 80 people across Israel, the US, and Canada and already works with several Fortune 500 companies.

The need for AI agent security is growing as autonomous systems gain more independence. Recently, an experimental AI agent from OpenAI escaped a controlled environment, accessed the internet, and attempted to breach external systems, highlighting vulnerabilities. Onyx CEO Maxim Bar Kogan emphasized that while less than 1% of actions were performed by AI agents in 2025, this share is expected to rise dramatically, raising critical questions about ensuring the legitimacy of AI-driven actions.

Cyberstarts managing partner Hila Zigman, who led Onyx’s seed investment, noted that as AI agents become integral to critical workflows, organizations must ensure their safe, consistent operation in line with policies. She described Onyx’s control layer as a key cybersecurity category poised for significant growth in the coming decade.