Defense tech funding in 2026: who funds dual-use startups
Defense and dual-use funding with every figure tied to its period: who writes the cheques, what traction means when your buyer is a government, and how CFIUS shapes your cap table.
Defense tech funding in 2026: who funds dual-use startups
Defense tech funding reached $49.9 billion across 966 deals in full-year 2025, per PitchBook, and Crunchbase News counted more than $14.6 billion raised in 2026 to date. The totals disagree because each dataset counts a different universe. Who backs you depends on whether you sell commercial first or government first.
Every sector guide tells you defense is hot. The useful part is different: a government buyer changes what an investor underwrites. In defense tech funding, procurement timelines, export control and cap-table ownership rules sit upstream of the term sheet and decide which funds can take the meeting at all.
As of August 2026, the most recent complete-period data on this page covers full-year 2025. PitchBook, CB Insights and Crunchbase publish on quarterly or annual cadences, not weekly ones, so every figure below is labelled with the period it describes. The freshest datapoint here is a 2026 year-to-date total, not a live number.
Defense tech VC by the numbers, and why the totals disagree
There is no single number for the size of this market, and any guide quoting one without naming the dataset is guessing.
| Figure | Period | What it counts | Source |
|---|---|---|---|
| $49.9B across 966 deals | Full-year 2025 | Defense tech VC, broadest cut | PitchBook |
| $11.9B across 225 deals | Q4 2025 | Same cut, single quarter | PitchBook |
| ~$9.5B across 265 rounds | Full-year 2025 | Global defense tech VC, narrower cut | PitchBook |
| ~$3.8B across 172 deals | Full-year 2024 | Same narrower cut, prior year | PitchBook |
| $1.5B raised, ~$6.0B projected | 2025, at report date | AI companies targeting defense applications | CB Insights |
| More than $14.6B | 2026 year to date | Defense tech, against a $9.6B record set in 2025 | Crunchbase News |
The gap between $49.9 billion and $9.6 billion for the same year is a definition problem, not an error. One cut counts the broadest defense tech universe, another counts a narrower set of rounds, a third counts only AI companies targeting defense applications. Put the dataset and the period next to any market size you show an investor.
Direction is what the sources agree on: on PitchBook's like-for-like cut, the 2024 and 2025 rows above are the same measurement, and the count of rounds rose with the capital.
What investors bought changed as well. PitchBook describes 2025 as an industrialization cycle in which investors favored production readiness, supply-chain resilience, deployable AI and scaled execution over exploratory prototypes, per PitchBook. If your 2026 raise narrative is a prototype and a vision, you are pitching into the previous cycle.
Who actually writes cheques in dual use startup funding
Four kinds of capital sit behind the label, and picking the wrong one costs you a quarter.
| Capital source | Named example | What it brings | Evidence |
|---|---|---|---|
| Government-linked strategic investor | In-Q-Tel | $250k to $500k equity-only, or $500k to $3M technology-development agreements | In-Q-Tel |
| Specialist national-security fund | Shield Capital | Judgment on commercial-first versus government-first sequencing | Shield Capital |
| Generalist fund with a defense thesis | Andreessen Horowitz, American Dynamism | Larger rounds for companies spanning defense and commercial markets | Andreessen Horowitz |
| Corporate venture arms | Corporate investors as a group | 28 rounds worth a combined $2B in 2025, 10.6% of deal activity | PitchBook |
Cheque size tells you what each source is for. The In-Q-Tel ranges above, per In-Q-Tel, are validation and access money. Treat them as a credential, not as your round.
Corporate money is now a structural part of the stack, at 10.6% of 2025 defense-tech deal activity, per PitchBook. Take the strategic cheque when it comes with a distribution path, and decline it when it comes with exclusivity.
What traction looks like for national security startups
National security startups get underwritten on a question no SaaS metric answers: can you survive the gap between a pilot and a funded program?
Say which way you sequence in the first line of the deck. Shield Capital says a commercial-first approach can apply to AI and cyber companies that build commercial traction before pursuing national-security use cases, while autonomy and space companies may approach the Defense Department or Intelligence Community first and pursue commercial customers later, per Shield Capital.
Government sales is a discipline, not a channel. Andreessen Horowitz presents the Department of Defense as a large and stable potential customer while treating government contracting as a specialized route founders must learn rather than ordinary enterprise sales, per Andreessen Horowitz. Budget that learning curve into runway, not into the appendix.
Partnerships with primes count as traction. CB Insights highlights defense-contractor and startup partnerships as a route to traction, including L3Harris combining electronic-warfare capabilities with Shield AI's autonomous-flight technology, per CB Insights. One signed integration with a prime beats three unfunded pilots on a slide.
✅ Good: "Two units fly the system weekly under a funded Phase II, and the program office has scheduled a production-readiness review." Names the user, the vehicle and the next funded step.
❌ Bad: "Strong interest from multiple DoD stakeholders and a large and growing pipeline." Interest is not a funded step, and a pipeline with no contract vehicle is not a forecast.
ITAR, CFIUS and your investor universe
Govtech defense investors are filtered by regulation before they are filtered by thesis.
CFIUS can apply to non-control investments in US businesses involving critical technology, critical infrastructure or sensitive data, and a fund's foreign limited partners or control rights can affect the analysis, per Cooley. That makes a fund's LP base your diligence item, not only theirs.
Ask two questions on the first call: where the fund's capital comes from, and what board or information rights it requires. Finding out during confirmatory diligence costs you the round, not the meeting.
Europe is not locked out. In-Q-Tel says it explicitly invests in international companies, including businesses based in the UK, EMEA, Asia-Pacific and Canada, per In-Q-Tel, so US national-security capital is reachable from a European entity, even though contracting and security requirements stay separate questions.
One honest gap: the datasets above report deal counts and capital totals, not stage-level defense startup valuations. Price off general seed valuation benchmarks and treat sector-specific valuation claims with no named dataset behind them as noise.
Turn the market read into a target list
Market data tells you what is normal. It does not tell you which funds are deploying into your stage, sector and geography right now, which is the only question that changes what you do on Monday.
Work the current tape in startup funding this quarter so your comps carry a period, then pull a live target list from active seed deep tech VCs, the closest match for most dual-use hardware and autonomy companies. If you sit on the hardware side, raising a seed round for a robotics startup goes deeper on the traction framing.
For the matching step itself, Causo maps your company to the investors most likely to fund it at your stage and drafts the outreach, which is the part a market report cannot do.
FAQ
Which VCs fund defense tech startups? Four groups, and they behave differently. Government-linked strategic investors such as In-Q-Tel write $250,000 to $500,000 equity-only cheques, or $500,000 to $3 million technology-development agreements, per In-Q-Tel. Specialist national-security funds such as Shield Capital and generalist firms with a defense thesis such as Andreessen Horowitz's American Dynamism practice cover the venture rounds. Corporate investors are the fourth group: they joined 28 defense-tech rounds worth $2 billion in 2025, or 10.6% of deal activity, per PitchBook.
How much are defense startups raising? PitchBook recorded $49.9 billion across 966 deals in full-year 2025, of which $11.9 billion across 225 deals landed in Q4 2025, per PitchBook. Crunchbase News reported more than $14.6 billion raised in 2026 year to date, above the $9.6 billion annual record it recorded for 2025, per Crunchbase News. The totals differ because each dataset counts a different universe, so name the source and period when you quote one. Individual 2025 rounds ran up to Saronic's $600 million Series C, per CB Insights.
What is dual-use technology? Dual-use technology is a product with a credible market in both commercial and national-security customers. Shield Capital says it is not limited by a traditional definition requiring a company to serve government and commercial customers simultaneously, and that timing plus credible potential for both markets can be sufficient, per Shield Capital. Sequencing splits by category: AI and cyber companies often build commercial traction first, while autonomy and space companies may approach the Defense Department or Intelligence Community first.
Can a European founder raise US defense money? Yes, with conditions. In-Q-Tel says it explicitly invests in international companies, including businesses based in the UK, EMEA, Asia-Pacific and Canada, per In-Q-Tel, so US national-security capital is reachable from a European entity. Contracting and security requirements are separate questions from investment. Screening runs the other way too: CFIUS can apply to non-control investments in US businesses involving critical technology, critical infrastructure or sensitive data, per Cooley.
Does ITAR prevent foreign investment in a defense startup? ITAR is an export-control regime covering defense articles and technical data, and it is a different question from who is allowed to own your equity. In the US the ownership screen is CFIUS, which can apply to non-control investments in businesses involving critical technology, critical infrastructure or sensitive data, and where a fund's foreign limited partners or control rights can affect the analysis, per Cooley. Foreign capital is not automatically blocked, but foreign LPs, board seats and information rights all change the analysis. Ask each fund about its LP base and required rights on the first call.
Related on the hub
- Startup offer letter template (with equity grant) — Startup offer letter template with the equity section that actually converts candidates: options, st…
- The H1 2026 AI startup funding report — Related fundraising basics guide.
- Raising a seed round for an AI agent startup in 2026 — Related fundraising basics guide.
- AI Startup Funding This Quarter: Deals, Sizes, Valuations — Related fundraising basics guide.
Run this raise inside Causo.
Match to the best-fit partner at 1,000+ funds, draft a hyper-specific email, and send from your own inbox, in one place.