Cybersecurity and privacy startups raised $10.6 billion in the first half of 2026, which Crunchbase News describes as roughly in line with recent comparable periods. Read only that line and the market looks stable.
The quarterly split tells a different story. Q2 brought in $4.4 billion — down about 30% on Q1, and down about 30% on Q2 2025. Round counts fell by a similar margin.
A half-year total that holds while the second quarter drops 30% means the first quarter was carrying it. And the reason Q2 didn't fall further is that eight rounds of $100 million or more landed inside it.
Where the money actually went
| Company | Raised | Valuation |
|---|---|---|
| Cyera | $600m | $12bn |
| NinjaOne | $400m+ (Series C extension) | $12.3bn |
| Dream | $260m | $3bn |
More recently, consumer privacy and security platform Cloaked took a $375 million Series B, and Tenex.AI, selling AI-driven security services, raised a $250 million Series B.
Note what those have in common. They are late-stage cheques into companies that already have distribution, or into the specific intersection of security and AI. None of them is evidence that it got easier to raise a first round.
The number founders should read
Total capital is the wrong metric if you are the one fundraising. You cannot access a market total — you access a cheque, and cheques come from deals.
Deal count falling roughly in line with dollars, while eight companies absorb nine figures each, describes a barbell: a small number of very large rounds at the top, and a thinner, harder market underneath. The average conceals both ends.
This is also why "cybersecurity funding is at historically high levels" and "it is harder to raise than last year" can both be true at once, and usually are.
The AI comparison is the context
Crunchbase's read is that the Q2 dip is moderate rather than alarming — a decline against unusually strong prior quarters, not a warning sign.
The more structural point in the same analysis is that security is no longer where investor attention concentrates. AI is. That matters less for the capital available and more for the terms and the pace: sectors that aren't the current thesis get diligenced longer and priced harder.
Which is roughly what the Tenex.AI round suggests founders have noticed — the security companies raising largest are the ones that can credibly describe themselves as AI companies.
If you're raising into this
- Model on deal count, not sector totals. The headline number includes rounds you were never eligible for.
- Assume a longer process. Fewer deals at similar dollars means more time per deal, not less.
- Be precise about the AI claim. "AI-powered" is now a filter both ways: absent, you look dated; overstated, you get tested on it in diligence.
- Late-stage strength is not early-stage weather. A $600m round at a $12bn valuation says something about Cyera. It says nothing about the market for your seed.
The money is still in the sector. It is just arriving in fewer, larger pieces, which is a different market to raise into even when the annual chart looks flat.
