Israel’s cybersecurity startup Onyx Security emerges from stealth with $40M to secure the next wave of AI agents

Artificial intelligence agents are moving from research labs into corporate systems at an astonishing pace. Companies now let software agents read data, make decisions, and execute tasks across internal platforms. That shift creates a new security problem that few organizations are ready to handle. A new Israeli cybersecurity startup thinks it has the answer.

Onyx Security officially launched from stealth on Thursday with $40 million in initial funding and a clear mission: to secure the growing army of AI agents operating inside modern enterprises. The launch signals continued investor confidence in Israel’s tech ecosystem during a tense geopolitical moment.

Onyx spent the past year and a half operating quietly in stealth. During that period, the company built a platform focused on controlling and securing AI agents that interact with sensitive corporate systems. The startup says several Fortune 500 companies are already testing the technology.

Onyx Security Launches From Stealth With $40M to Secure the Next Wave of AI Agents

The timing reflects a growing concern across corporate security teams. AI agents can now access internal tools, financial systems, customer databases, and production environments. That level of access introduces new risks if an agent behaves unpredictably or becomes compromised.

“Every enterprise is becoming an agent operator — whether they planned to or not,” said Maxim Bar Kogan, co-founder and CEO of Onyx, who served in Israel’s elite Unit 8200 military intelligence unit.

Agents are already performing tasks once reserved for employees. They analyze data, interact with internal APIs, execute workflows, and automate decision-making. That shift raises a critical question for security leaders.

“Agents are given access to the most critical ‌systems ⁠in the enterprise, but what are our guarantees they will not make serious mistakes or get compromised?” Bar Kogan said. “The safe adoption of AI agents requires security from attacks, as well as ensuring agents don’t make critical ⁠mistakes.”

The startup’s platform focuses on monitoring AI agents’ behavior, managing their permissions, and preventing unauthorized actions across enterprise systems. The idea mirrors identity and access management tools used for human employees, extended to autonomous software agents.

Investors appear convinced the opportunity is real. The $40 million round was led by venture firms Conviction Partners and Cyberstarts, along with a group of angel investors. Onyx says the capital will fund product development, new AI models, and expansion of engineering teams.

The launch arrives during a strong year for Israel’s startup ecosystem. The country remains one of the world’s largest tech hubs, with the sector responsible for roughly 20% of GDP, 15% of jobs, and more than half of exports, Reuters reported. Cybersecurity continues to attract strong investor interest.

Funding numbers reflect that momentum. Israeli startups raised about $16 billion in 2025, up from $12.2 billion in 2024, showing renewed investor confidence in the country’s technology sector.

A major deal closed just days before Onyx stepped out of stealth. Google’s parent company, Alphabet, completed its $32 billion acquisition of Israeli cybersecurity firm Wiz, one of the largest tech exits in Israel’s history. The deal reinforced global demand for security platforms built in the country.

Onyx is betting that the next wave of cybersecurity will revolve around AI agents rather than traditional software vulnerabilities. As companies grant autonomous systems greater authority within their infrastructure, controlling those systems becomes a new priority.

Security teams once focused on defending networks from outside attackers. The challenge now extends to managing what autonomous software does once it gains access to internal systems.

Onyx wants to be the company standing at that gate.

Onyx Security team