Armadin Raises a Record $189.9M in Seed and Series A Funding to Deploy an Autonomous Agentic Attacker Swarm — The Largest Combined Early-Stage Round in Cybersecurity History
Led by Accel with participation from Google Ventures, Kleiner Perkins, Menlo Ventures, In-Q-Tel, 8VC, and Ballistic Ventures, Armadin's record round backs an AI-native platform that fields specialised AI agents trained to the standards of a world-class red team — reasoning, planning, and adapting like advanced human threat actors at machine speed, to identify and prove exploitable risk before AI-powered attackers do.
5 min read
Armadin, an AI-native cybersecurity company, has raised $189.9 million in combined Seed and Series A funding — the largest combined early-stage funding round in cybersecurity history. The round was led by Accel, with participation from Google Ventures, Kleiner Perkins, Menlo Ventures, In-Q-Tel, and follow-on investment from 8VC and Ballistic Ventures. The company is led by Kevin Mandia — one of the most widely recognised names in cybersecurity incident response — and is building a platform that deploys an autonomous agentic attacker swarm to prepare organisations for the speed and scale of AI-driven threats.
The scale of the funding — and the calibre of the investor syndicate, which includes In-Q-Tel, the strategic investment arm associated with the US intelligence community — reflects both the urgency of the threat environment Armadin is addressing and the conviction that autonomous offensive AI capabilities are becoming a foundational requirement for enterprise and national security defence. The round is not just a company milestone: it is a signal about where the security industry believes the next phase of the arms race is heading.
"The AI shift is changing cybersecurity more rapidly than any transition in history. In a world of machine-speed attacks, defence must become autonomous. You cannot have a human in the loop for every defence decision and expect to win. We are building the most formidable offence to give organisations the greatest defence. It's important to national security."
— Kevin Mandia, CEO, Armadin
The Hyperattack Gap — Why Human-Led Defence Can No Longer Keep Pace
The threat model Armadin is built around is the emergence of what it calls Hyperattacks — sophisticated, multi-modal attack campaigns that move at machine speed, combining AI-driven reconnaissance, vulnerability exploitation, lateral movement, and data exfiltration into coordinated operations that unfold faster than any human security team can detect, triage, and respond to. Traditional human-led defences operate on timescales measured in hours and days; Hyperattacks operate in minutes and seconds. The gap between those two timescales is the structural vulnerability that Armadin is designed to close.
Armadin's approach to closing that gap is conceptually direct: if AI-powered attackers are the threat, the most effective defence is an AI-powered attacker that continuously tests your own defences before the adversary does. By deploying the ultimate attacker against its own customers' networks — continuously, autonomously, at machine speed — Armadin identifies and proves exploitable risk in real time, giving security teams and executive leadership decision-grade proof of what can actually be exploited rather than theoretical vulnerability scores or penetration test reports that are outdated by the time they are delivered.
The Agentic Attacker Swarm — AI Agents That Reason, Plan, and Adapt Like Elite Human Threat Actors
Unlike conventional vulnerability scanners or automated penetration testing tools that follow predefined scripts or rule sets, Armadin's platform features specialised AI agents leveraging custom models in an agentic attacker swarm. These agents continuously reason, plan, and adapt — mimicking the behaviour and decision-making of the most advanced human threat actors, including nation-state adversaries. The distinction between scripted automation and genuine agentic reasoning matters enormously in offensive security: the most dangerous attacks are not the ones that follow known patterns, but the ones that adapt to the specific topology, defences, and opportunities of the target environment in real time.
"The most honest measure of security has always been the offensive lens. At Armadin, we are taking decades of human-led red teaming expertise and reinforcing it into AI models. These models are learning our tactics and techniques and are outpacing our human operators at every turn."
— Evan Peña, Founder and Chief Offensive Security Officer, Armadin
"Security expertise is a constrained resource that organisations never have enough of in the moments when it matters most. Before Armadin, you could not put a nation-state level adversary inside every network 24/7. We've built the ultimate attacker — it doesn't just follow a script, it reasons and learns as it swarms your defences. We train our models and build agents to the standards of a world-class red team with safety at the foundation and unleash them to identify exploitable risk at machine speed. We believe that this is the only way to prepare for the coming wave of AI Hyperattacks."
— Travis Lanham, Founder and CTO, Armadin
Investor Conviction — What the Syndicate Signals About the Market
The investor syndicate behind Armadin is notable for both its depth and its diversity. Accel led the round; Kleiner Perkins, Menlo Ventures, and Google Ventures represent some of the most established names in enterprise technology venture investment; Ballistic Ventures is a specialist cybersecurity fund with deep sector expertise; and In-Q-Tel's participation signals direct alignment between Armadin's technology and national security priorities. Together, the syndicate reflects a shared conviction that autonomous offensive AI capabilities are not a future category — they are an immediate and urgent requirement.
"Armadin is the first company we've seen that truly weaponises the attacker's perspective to build a more resilient defence. By combining Kevin's unrivalled operational experience with a generational AI engineering team, Armadin is delivering the autonomous, comprehensive system of record for an enterprise's security posture that boards and CISOs have been demanding for years."
— Ping Li, Partner, Accel
Armadin's founding team combines elite red teaming expertise with AI research and engineering capability — a rare pairing that is central to the platform's differentiation. The offensive security knowledge encoded into Armadin's models reflects decades of real-world adversarial experience, including the tactics, techniques, and procedures used by the most sophisticated threat actors. That expertise, reinforced into AI models that learn and adapt continuously, is what distinguishes Armadin's agentic attacker swarm from generic automated security testing tools. More information is available at armadin.ai.
Key Takeaways
- Armadin has raised $189.9M in combined Seed and Series A funding — the largest early-stage round in cybersecurity history — led by Accel with participation from Google Ventures, Kleiner Perkins, Menlo Ventures, In-Q-Tel, 8VC, and Ballistic Ventures, reflecting investor conviction that autonomous offensive AI capabilities are an immediate and urgent enterprise and national security requirement
- The platform addresses the "Hyperattack Gap" — the structural mismatch between AI-powered attacks that operate at machine speed and human-led defences that operate on timescales measured in hours and days — by deploying an autonomous agentic attacker swarm that continuously tests enterprise defences at the same speed and sophistication as the adversaries organisations are facing
- Unlike scripted vulnerability scanners, Armadin's specialised AI agents reason, plan, and adapt like advanced human threat actors — including nation-state adversaries — continuously identifying and proving exploitable risk in real time and delivering decision-grade proof to CEOs and Boards rather than theoretical vulnerability scores or outdated penetration test reports
- The platform's models are trained to world-class red team standards, encoding decades of human offensive security expertise and real-world adversarial tactics — with AI agents that are already outpacing their human operator counterparts, and a safety-at-foundation design that ensures the attacker swarm operates within defined parameters at all times
- In-Q-Tel's participation in the round signals direct alignment with national security priorities — reflecting that autonomous AI-native offensive security capabilities are now considered strategically critical infrastructure by the US intelligence community, not just a commercial enterprise security product category