Mandia Launches Armadin With Record $190M Seed

Kevin Mandia doesn't do small comebacks. The man who built Mandiant into the world's most feared incident response firm, then sold it to Google for $5.4 billion in 2022, just raised $189.9 million in combined seed and Series A funding for his new venture — an AI-native cybersecurity startup called Armadin. It's the largest early-stage funding round in cybersecurity history. And it tells you everything about where security is headed.

The Thesis: Humans Are Too Slow

Mandia's pitch is blunt and uncomfortable: human defenders can't keep up anymore. AI-powered attackers are executing in minutes what used to take days. He calls them "hyperattacks" — sophisticated, multi-modal campaigns operating at machine speed, orchestrated by AI agents that never sleep, never forget, and never hesitate.

His answer? Fight fire with fire. Armadin is building autonomous AI agents that function as an "attacker swarm," continuously reasoning, planning, and adapting to simulate advanced threat actors before the real ones show up. Think of it as a relentless red team that runs 24/7, probing your entire attack surface at the same speed adversaries would.

"In a world of machine-speed attacks, defense must become autonomous. Human intervention is too slow."
— Kevin Mandia, CEO of Armadin

That's not marketing fluff from a first-time founder. That's a conviction forged across decades of responding to the worst breaches on the planet — from the SolarWinds attack to countless nation-state intrusions. When Mandia says human-speed defense is dead, the market listens.

The Money: A Who's Who of Smart Capital

The $189.9 million round was led by Accel, with a murderer's row of co-investors: GV (Google Ventures), Kleiner Perkins, Menlo Ventures, 8VC, Ballistic Ventures, and — notably — In-Q-Tel, the CIA's venture arm.

Let that last one sink in. When the intelligence community's investment vehicle writes a check into your seed round, it's not just about returns. It's a signal that the national security apparatus considers your technology strategically important. Armadin isn't just building a product — it's building infrastructure that the U.S. government wants to exist.

The presence of GV is equally telling. Google acquired Mandiant and folded it into Google Cloud Security. Now Google's venture arm is backing Mandia's next act. Either they believe in the founder so much they'll fund him regardless, or they see Armadin as complementary to what they're building. Probably both.

The Team: Ex-Google, Ex-Mandiant, All Killer

Mandia didn't just raise capital — he assembled a founding team that reads like a cybersecurity all-star roster:

• Travis Lanham (CTO) — Former Google Cloud Security principal engineer
• Evan Peña (Chief Offensive Security Officer) — Former Mandiant executive
• David Slater (Chief Architect) — Former Google SecOps engineer

This is a team that has built and operated security at the highest levels of both the private sector and government-adjacent work. They know what breaks, where it breaks, and why traditional approaches can't scale to meet AI-driven threats.

The company was incorporated in September 2025, closed its seed round by October, and had Series A capital locked down by late December. In six months, Armadin has hired over 60 employees and is already working with Fortune 100 companies. That's not startup speed. That's escape velocity.

Why This Matters: The AI Security Arms Race Is Real

Let's be honest about the landscape. Every cybersecurity vendor is slapping "AI" on their marketing materials. Most of it is cosmetic — a chatbot here, some ML-assisted alerting there. Armadin is making a fundamentally different bet: security itself must be AI-native from the ground up, not AI-augmented on top of legacy architectures.

The distinction matters enormously. Augmented tools still require humans in the loop for critical decisions, creating bottlenecks that attackers can exploit by simply moving faster than SOC analysts can respond. A truly autonomous system — one that reasons, adapts, and acts at machine speed — doesn't have that bottleneck.

Of course, autonomous security agents come with their own risks. False positives at machine speed could be catastrophic. An AI red team that's too aggressive could disrupt production systems. The governance questions around autonomous offense-simulating AI are thorny and largely unresolved. Mandia's team will need to solve these problems in production, not just in pitch decks.

The Competitive Landscape

Armadin enters a market that's simultaneously crowded and starving. Crowded with legacy vendors bolting on AI features. Starving for genuinely new approaches. Companies like CrowdStrike, SentinelOne, and Palo Alto Networks dominate the market but were all built in the pre-LLM era. Startups like Armadin are betting that the AI inflection point demands entirely new architectures — not patches on old ones.

Mandia's track record gives Armadin an enormous unfair advantage. He doesn't need to convince CISOs that he understands real threats. He literally wrote the book. The trust capital alone is worth more than the $189.9 million in the bank.

The Bigger Picture

This raise is a bellwether. When the most credible founder in cybersecurity raises the largest early-stage round in the sector's history, built entirely around autonomous AI agents, it's not just a company launch — it's a market declaration. The era of human-centric security operations is ending. The era of AI-native security infrastructure is beginning.

The question isn't whether autonomous security agents will become the standard. It's who will build the ones that work. With Mandia's credibility, this team, this capital stack, and the backing of the U.S. intelligence community, Armadin is making an aggressive early claim on that future.

Now they just have to build it. Given who's at the helm, I wouldn't bet against them.