Last Updated: 30 July 2026
Onyx Security, a startup that builds a monitoring and control platform for enterprise AI agents, has raised $113 million in Series B funding. Co-founded by CEO Maxim Bar Kogan and Chief AI Officer Gil Elbaz, the company sells what it calls a "secure AI control plane," a system that lets security and governance teams watch over the AI agents operating inside their organizations and enforce rules on what those agents can do.
The round comes at a time when companies are adopting AI agents far faster than they can secure them. AI agents are pieces of software that can make decisions, access company systems, and take actions on their own, without a human approving each step. According to MarketsandMarkets, the agentic AI security market is projected to grow from $1.65 billion in 2026 to $13.52 billion by 2032, a compound annual growth rate of 42%. A Gravitee survey of 750 senior technology leaders found that enterprise AI agent fleets roughly doubled between December 2025 and April 2026, while security coverage barely moved.
Bessemer Venture Partners led the round, with participation from Cyberstarts, TCV, Conviction, FirstMark, Vintage Investment Partners, QuantumLight, and G Squared. The round brings Onyx's total funding to $153 million since the company was founded two years ago.
Most cybersecurity tools were designed for a world where humans are the ones using company systems. Employees log in, open files, send emails, and security software tracks what they do. AI agents work differently. They can browse the web, write code, pull data from databases, and carry out tasks across multiple systems, often without anyone watching in real time. When something goes wrong, like a data leak or an unauthorized action, it can be very difficult to figure out which agent did what, and why.
That problem is getting bigger as AI models become more capable. Some recent models reached such advanced levels of performance that they were initially considered too risky for public release. As these models power enterprise agents, employees are giving them access to critical systems in banking, energy, insurance, and technology, and tasking them with increasingly complex work. The agents can act, but as Onyx has noted, they are not accountable for their actions.
"AI agents are becoming an important part of how modern companies operate, but enterprises need trust, control, and accountability before they can adopt them at scale. Onyx is building the control layer that makes this possible. Maxim and Gil have identified one of the most urgent challenges in enterprise AI, and we believe that Onyx is positioned to define this category."
Amit Karp, Partner at Bessemer Venture Partners
The new capital will go toward two priorities: training the next generation of Onyx's proprietary AI models and scaling the company's enterprise go-to-market in the United States and beyond. The proprietary models sit at the core of the platform. They follow each step of an AI agent's reasoning and can step in to correct unintended or malicious agent actions in real time.
Revenue has quadrupled since Onyx came out of stealth four months ago. In June, Anthropic announced an integration with Onyx to help secure enterprise AI adoption, joining dozens of other AI providers already integrated with the platform, including OpenAI, Amazon, Microsoft, and Google. According to the company's blog, the platform secures more than 1.1 million agents across enterprise deployments and has analyzed over 66 million sessions.
"As AI agents become embedded in critical business workflows, enterprises need a way to ensure they operate safely, predictably, and within policy. Onyx is building the control layer that makes enterprise AI adoption possible at scale. We believe this will become one of the defining security categories of the coming decade."
Hila Zigman, General Partner at Cyberstarts
Maxim Bar Kogan and Gil Elbaz founded Onyx Security two years ago. Bar Kogan serves as CEO; Elbaz as Chief AI Officer. The company is headquartered in Tel Aviv with a U.S. office in New York. Bar Kogan served in a technology unit in the Israel Defense Forces and won the CodeGuru Israel competition in 2008. The board of directors includes partners from Bessemer Venture Partners, Cyberstarts, and Conviction, along with executives from cybersecurity firm Armis.
Think of Onyx as a control room for AI agents. Whether a company buys agents from outside vendors or builds them internally, Onyx sits in between, watching what each agent does. It sees every prompt sent to an agent, every response it gives, and every action it takes. If an agent tries to access restricted data or run an unauthorized command, Onyx can block it. The platform covers agents running across SaaS applications, cloud environments, employee devices, and code repositories.
Onyx counts dozens of Fortune 500 companies among its customers, spanning banking, technology, insurance, energy, and other industries. Uros Solar, Head of Security Engineering at Revolut, has said publicly that Onyx provides the visibility and control Revolut needs to use coding agents safely across multiple service providers.
Bessemer Venture Partners, which led the round, has backed more than 420 companies and counts over 150 IPOs. Its portfolio includes Anthropic, Shopify, and LinkedIn, and the firm maintains a dedicated cybersecurity investment practice. Cyberstarts, which has been involved with Onyx from earlier stages, is a cybersecurity focused fund founded in 2018. Its portfolio companies represent over 50% of the worldwide private cybersecurity market capitalization, according to the firm, with investments in Wiz, Cyera, and Island among others.
Other participants in the round include TCV, Conviction, FirstMark, Vintage Investment Partners, QuantumLight, and G Squared. The pace is worth noting: the company reached $153 million in total funding just two years after its founding, with this Series B arriving only four months after its public launch.
