Anthropic Just Integrated This AI Security Startup Into Its Enterprise Stack. It Then Raised $113 Million.

Onyx Security launched four months ago. Since then revenue has grown fourfold, Anthropic has integrated its technology, and the company just closed $113 million in Series B funding.

The New York and Tel Aviv-based startup announced its Series B led by Bessemer Venture Partners, with participation from Cyberstarts, TCV, Conviction, FirstMark, Vintage, QuantumLight, and G Squared. The round values Onyx at $640 million and brings total funding to $153 million since founding two years ago.

Onyx builds what it calls an AI control plane — a unified layer that gives security, governance, and infrastructure teams visibility and control over AI agents operating across the enterprise. The problem it is solving is specific: as companies deploy AI agents into production workflows, those agents have access to sensitive systems — CRM data, code repositories, financial records, internal documents. The traditional security stack cannot monitor or govern what an AI agent does with that access at runtime.

The Anthropic integration is the clearest signal of where this market is heading. In June 2026, Anthropic announced it had integrated Onyx's technology to help enterprise customers adopt Claude securely. When an AI model company integrates a third-party security layer into its enterprise deployment stack, it is making a statement: agentic AI in the enterprise requires a purpose-built governance layer, and the AI providers themselves are not building it.

For context, Zenity raised $125 million for a similar thesis just weeks earlier. Two companies in the AI agent security category closing nine-figure rounds within a month of each other is a signal that this is now a confirmed market category, not a niche.

What this means for founders

Onyx's revenue trajectory — 4x growth in the four months since coming out of stealth — is the number that matters most. Bessemer did not lead a $113 million round on potential. It led on demonstrated enterprise adoption moving fast. For founders building security, governance, or compliance infrastructure for AI deployments: the enterprise buyer's urgency has shifted from "we should think about this" to "we need this now." The sales cycle for AI governance tools in 2026 is compressing as AI agent deployments accelerate and CISOs realize their existing stack does not cover agentic risk.

Investors: Bessemer Venture Partners, Cyberstarts, TCV, Conviction, FirstMark Round: Series B — $113M Valuation:$640M | Revenue growth: 4x since stealth launch (4 months)

Powered by AngelLinx Intelligence. Updated daily. angellinx.ai