How the autonomous pentesting pioneer plans to scale after its biggest funding round yet

Introduction

San Francisco‑based Horizon3.ai has secured a $73 million Series C round that values the four‑year‑old cybersecurity startup at roughly $500 million, according to people familiar with the terms. The raise, led by Growth Partners Capital with participation from SignalFire, AllegisCyber, and private investors including CrowdStrike co‑founder Dmitri Alperovitch, comes amid a record year for ransomware and a tightening venture market. For Horizon3.ai, it is both validation of its autonomous pentesting platform—NodeZero—and a war chest to pursue an aggressive go‑to‑market strategy.

1. What Horizon3.ai Does

Founded in 2019 by former U.S. Special Operations cyber officer Snehal Antani and ex‑NSA engineer Anthony Pillitiere, Horizon3.ai developed NodeZero, a software‑as‑a‑service platform that behaves like an “automated red team.” Using continuous reconnaissance, privilege escalation, and exploit chaining, NodeZero identifies attack paths from asset discovery to domain compromise—then generates remediation guidance. Unlike traditional penetration‑testing consultancies that operate quarterly or annually, NodeZero can be run as often as customers like, with no agents or credentialed access required.

2. Inside the $73 Million Series C

Growth Partners Capital took the lead with a $40 million check; existing investors SignalFire and AllegisCyber re‑upped for another $25 million combined, while strategic angels covered the balance. The round includes a small secondary component for early employees and brings total funding to $120 million. Horizon3.ai declined to disclose revenue, but CEO Antani says annual recurring revenue has tripled for three consecutive years and surpassed $40 million in Q1 2025.

3. Why Investors Are Betting Big

• Market Pull — Gartner forecasts the attack‑surface‑management and breach‑and‑attack‑simulation segments will reach $3 billion by 2027, a 27 percent CAGR.

• Differentiated Tech — NodeZero uses a proprietary chain‑analysis engine that scored 97 percent exploit success in MITRE ATT&CK alignment tests, according to a February 2025 SANS review.

• Federal Momentum — The company recently achieved FedRAMP In Process status, opening doors to U.S. civilian agencies and potentially DoD contracts.

• Capital Efficiency — Horizon3.ai claims a burn multiple below 1.5, rare among Series C startups.

4. How the Money Will Be Used

Antani says 40 percent of proceeds will expand R&D, with priorities including an AI‑driven exploit developer and a cloud misconfiguration module. Another 35 percent funds international sales hubs in London and Singapore. The balance supports compliance and channel‑partner programs. A modest inorganic strategy—acquiring boutique firms with specialist talent—is “on the radar” but not immediate, the CEO notes.

5. Competitive Landscape

The autonomous‑testing field is crowded: Israel’s Pentera raised $150 million at a $1 billion valuation in 2024; IBM bought Randori for an undisclosed sum; and Rapid7 acquired InsightOps to flesh out its offensive‑security suite. Horizon3.ai argues its agentless architecture and pay‑as‑you‑test pricing create a moat, but the big cloud providers are looming—Microsoft’s Security Copilot hinted at future breach‑simulation features in its latest roadmap.

6. Risks and Challenges

• **Customer Education** — Many CISOs still view pentesting as a compliance checkbox, not a continuous process.

• **Regulatory Scrutiny** — Autonomous exploit tools can trigger export‑control questions, especially with dual‑use capabilities.

• **Talent War** — Offensive‑security engineers command six‑figure salaries; scaling without diluting culture will be tricky.

• **Macro Headwinds** — If budgets tighten, “nice‑to‑have” tools may face deferral, though ransomware headlines may offset this.

Conclusion

By locking in $73 million, Horizon3.ai gains more than cash: it secures a vote of confidence that autonomous offensive security has crossed from experiment to enterprise must‑have. Yet the coming year will test whether the startup can convert market buzz into durable, large‑scale deployments—and whether continuous pentesting can truly become as routine as patching. For now, the horizon looks promising, if crowded.