AI Security Startup Shatters Records with $190M Raise

$190 million. That’s what Mandiant’s founder just raised for his new AI security startup. Out of the gate. Seed and Series A combined.

Kevin Mandia closed the round Tuesday. The AI security startup—called Armadin—claims it’s a record for security companies at this stage. Accel led. GV, Kleiner Perkins, Menlo Ventures, 8VC, Ballistic Ventures, and the CIA’s venture arm In-Q-Tel all participated.

No valuation disclosed. Smart.

When I ran TaskFlow, we kept our metrics close until we had leverage. Mandia knows the playbook.

Mandia built Mandiant from zero in 2004. Sold it to Google for $5.4 billion in 2022. Then became a VC at Ballistic Ventures—the security fund co-founded by Ted Schlein, formerly of Kleiner Perkins.

Now he’s back building. Same category. Different threat.

**Why This AI Security Startup Matters**

Autonomous agents. That’s what Armadin builds.

Software that learns threats and responds without humans in the loop. The pitch: AI hackers are coming. Security teams need their own AI armies to fight back.

Mandia told CNBC he believes autonomous AI attackers will hit soon. They’ll complete attacks in minutes that used to take days.

“When you have AI on offense, what you are going to get is a technology that can think, can learn, can adapt,” he warned.

Security researchers and government agencies echo this. AI already lowers the bar for sophisticated attacks. Bad actors get smarter tools. White hats need equivalent firepower.

That’s the AI security startup thesis.

**The Team**

Mandia brought former Mandiant and Google talent. Co-founders include Travis Lanham, former Google Cloud Security principal engineer. Evan Peña, former Mandiant exec. David Slater, former Google SecOps engineer.

Google DNA runs through this team. Makes sense given Mandia’s exit.

**The Numbers**

Company claims $189.9 million is a record for security startups this early. They’re right—mostly.

1Password raised $200 million Series A in 2019. But the company was 14 years old. OneTrust raised $200 million same year. Three years old and already scaling.

Armadin did this fresh. No prior revenue to point at. Pure bet on Mandia’s track record and the AI threat thesis.

That’s what $5.4 billion exits buy you. Credibility.

**What They’re Building**

Autonomous cybersecurity agents designed to counter AI-powered attacks. Think of it as agentic armies for the good guys.

Black hats get AI tools that adapt and learn. White hats need the same. Armadin aims to arm security teams with automated defenders that match attacker sophistication.

Specifics on product remain thin. Typical for this stage. Most pre-launch security companies keep capabilities quiet until customers deploy.

Smart move. Don’t telegraph your defenses to attackers.

**The Investor Bet**

VCs bet on three things here:

Mandia’s execution. He built and sold Mandiant for $5.4 billion. Pattern recognition matters. Second-time founders with billion-dollar exits get blank checks.

Market timing. AI security concerns accelerate daily. Enterprises face threats they can’t manually defend against. Autonomous defense becomes table stakes.

Team pedigree. Google and Mandiant alumni with deep security expertise. Hard to assemble this talent without a proven founder.

Raising $190 million isn’t success. It’s a liability. Now Armadin must execute against massive expectations.

Most startups that raise this early at this scale stumble. Burn climbs. Pressure mounts. Execution gaps widen.

Question is whether Mandia’s second act matches his first.

**Market Context**

Cybersecurity funding has cooled from 2021 peaks. Most security startups raise $10M-$30M Series A rounds. $190 million combined seed and A is an outlier.

But AI security represents a category shift. Enterprises can’t defend against autonomous attackers with manual processes. They’ll pay for automated defense.

TAM is massive. Global cybersecurity spending exceeds $200 billion annually. AI-native security could capture 10-20% of that as threats evolve.

Competition will follow. Every security vendor will add “AI agents” to their pitch deck. Armadin’s advantage: purpose-built from day one, not bolted onto legacy architecture.

**What’s Next**

Product launch timeline unclear. Most security startups take 12-18 months from funding to general availability.

Armadin likely already has design partners—enterprises testing early versions under NDA. Security companies don’t raise $190 million without customer validation.

First revenue likely comes from Fortune 500 security teams facing sophisticated threats. Pricing will reflect enterprise budgets—probably six figures annually minimum.

Next milestone: first major customer logo. When a bank or tech giant deploys Armadin’s agents, competitors will pay attention.

Mandia built Mandiant into the gold standard for incident response. Now he’s betting autonomous agents become the standard for defense.

Execution beats ideas. Every time. Mandia proved he can execute once. Now he does it again—at bigger scale, faster pace, higher stakes.