Onyx Security, an Israeli cyber startup, has announced a substantial $113 million Series B funding round to bolster its pioneering work in securing and governing autonomous AI agents within enterprise settings. The investment was spearheaded by Bessemer Venture Partners, with additional participation from prominent investors including Cyberstarts, TCV, Conviction, FirstMark, Vintage Investment Partners, QuantumLight, and G Squared. This latest round brings Onyx Security's total funding to an impressive $153 million, achieving an estimated valuation of $640 million.
The company, founded in 2024 by Maxim Bar Kogan and Gil Elbaz, is addressing a rapidly emerging challenge: the proliferation of AI agents that increasingly perform complex tasks and make decisions with limited human intervention. As enterprises delegate more workflows to these intelligent agents, the demand for robust control and security mechanisms becomes paramount. Onyx Security's platform, dubbed the 'Secure AI Control Plane,' is designed to discover 'shadow AI' across an enterprise, monitor agent actions, and apply real-time security, governance, and compliance controls.
The new capital will be strategically deployed to accelerate the development of Onyx's next-generation proprietary AI models, which are crucial for monitoring agent intent and reasoning. Furthermore, the funding will fuel the company's expansion of its sales, business development, and marketing efforts, particularly in the United States and other international markets. With AI agents projected to perform a significant majority of enterprise actions within the next year, Onyx Security positions itself as a critical enabler for safe and responsible AI adoption, especially for sectors handling critical infrastructure like energy, financial services, and healthcare where potential errors could have severe consequences.
Cybersecurity startup Exein raised $270 million in Series C funding at a $1.7 billion valuation, becoming Europe's most valuable cybersecurity scaleup. The Rome-based firm's Photon technology provides kernel-level runtime protection for over two billion connected machines including robotics, drones, and autonomous vehicles.
US President Donald Trump announced plans on September 19 to establish a dedicated 'AI Force' within the military and appoint an 'AI Czar' to oversee national artificial intelligence policy.
Cybersecurity researchers from Indian startup Hacktron AI used Anthropic's Claude model to breach OpenAI's internal systems during an authorized security audit. The team demonstrated access to employee accounts and internal code repositories in under 72 hours, earning a $6,500 bug bounty.
