Defense Venture Capital

Global VC investment in defense-related companies sat in a fairly flat band for years before breaking upward. PitchBook's full-year deal value series tells the story cleanly:

At the current run rate, 2026 deal value is on pace to meet or exceed 2025's full-year total by year-end, even as deal count continues to soften. Deal count peaked in 2021 and has never returned to that level, which is consistent with the barbelling pattern above: money is concentrating in fewer, larger bets rather than spreading across more companies.

Other data providers, using different company universes, confirm the direction even where the absolute numbers diverge:

• CNAS: ~$40B into defense and dual-use companies in 2025, more than double 2024.
• McKinsey/Reagan Institute NSIB Report Card: $56B in 2025 (+83% YoY), or $96B including OpenAI's defense-adjacent contracts (+158%).
• Crunchbase (narrower “military, national security, law enforcement” taxonomy): $9.6B for full-year 2025, already exceeded by May 2026 at $14.6B.
• Bain (deals >$10M only): >$16B for 2025.

None of these figures are wrong; they measure different populations of companies. See the methodology note at the bottom of this memo before citing a specific total externally.

advanced computing & software leads at $17.8B (128 deals, +457% YoY), autonomous systems follows at $12.3B (175 deals, +149%), then sensing/connectivity/security at $7.9B (142 deals).

AI = 82% of the sector. The Reagan Institute/McKinsey NSIB Report Card puts AI and AI-adjacent investment at 82% of total 2025 defense tech funding, up from 72% in 2024. Among rounds over $1B, 56% went to AI-related companies in 2025, versus 37% in 2024.

Q2 2026's ten largest deals reinforce both trends simultaneously, heavy AI/autonomy weighting at the top (Anduril, Groq) alongside real dollars flowing to nominally non-defense categories that still register in the defense tech taxonomy:

See supporting chart: PitchBook's Q2 2026 quarterly deal activity chart (2021-Q2 2026), which shows the barbell effect visually, deal value climbing even as the deal-count line falls off from its 2025 peak.

Stage mix: later, bigger, fewer

PitchBook's stage data through Q3 2025: late-stage VC captured $14.7B YTD, venture-growth $20.4B, early-stage $4.0B (145 deals), seed just $729.1M (133 deals). The Reagan Institute's independent figures confirm the same rotation: seed-stage funding down 58% YoY in 2025 even as late-stage/growth funding grew 89%.

For anyone advising a founder on timing: the capital is rewarding companies with revenue and contract traction, not concept-stage bets. Bain's diligence checklist (credible end-state business model, pricing and sustainment plan, ability to retain culture through scale) describes what late-stage investors are now underwriting for.

Q2 2026

Source: Emerging Defense

Company | Stage | Amount | Valuation | Lead

Anduril | H | $5B | $61B | Thrive Capital, a16z

True Anomaly | D | $650M | $2.2B | Eclipse and Riot Ventures

Stark | A | $567M | $4B | Sequoia, Founders Fund

Hermeus | C | $350M | $1B | Khosla Ventures

Mach Industries | C | $300M | $1.8B | Ribbit Capital, Infinite Capital

Allen Control Systems | B | $200M | $2.2B | Smash Capital

Reliable Robotics | D | $160M | $991M | Nimble Partners

Skydio | F | $110M $4.4B | 8 Firms

Havoc AI | A | $100M | $900M | Cobalt Capital

Dominion Dynamics | A | $100M | - | Georgian

Twenty | B | $100M | $1B | Accel

Scout AI | A | $100M | $394M | Align Ventures, Draper Associates

~27 other firms < $100M

Q1 2026

Anduril $4B

Investors: Co-led by Thrive Capital and Andreessen Horowitz.

Saronic $1.75B

In March 2026, Saronic raised $1.75B for their Series D putting their post money valuation at $9.25B. The raise is to advance its mission of ensuring maritime superiority for the U.S. and its allies by delivering autonomous platforms at scale across defense and commercial sectors. An important component to Saronic's efforts is the development of Port Alpha - its next generation shipyard - alongside the rapid expansion of its existing production facilities in Louisiana and Texas

Investors: Led by Kleiner Perkins. New investors Advent International, Bessemer Venture Partners, DFJ Growth, BAM Elevate, as well as existing investors 8VC, Ca?einated Capital, Andreessen Horowitz, Elad Gil, and Franklin Templeton.

Hadrian $TBD

AI-powered autonomous defense manufacturing. New round to accelerate factory expansion, including the 270,000 sq ft Mesa, AZ "Factory 3" (4x prior throughput). Hadrian simultaneously announced a public-private partnership with the U.S. Navy on its "Factory of the Future" initiative, embedding autonomous CNC and additive manufacturing capabilities into naval supply chain programs. New additive manufacturing capacity expected online in 2026. Total raised since founding: nearly $500M+. Hadrian also acquired Datum Source, an AI-based manufacturing partner-matching platform founded by ex-SpaceX engineers.

Investors: T. Rowe Price (lead), General Catalyst, Founders Fund, Red Cell Partners, Fifth Down Capital, Narya, XYZ Venture Capital, a16z (existing)

Shield AI: $240M

Andreesen Horowitz (a16z)

Andreessen Horowitz announced a $15B new capital raise with $1.776B earmarked for its American Dynamism fund focused on defense, national security, and dual-use technology. Simultaneously announced a technology acceleration partnership with Booz Allen Ventures, giving a16z portfolio companies access to Booz Allen's government market relationships in AI, autonomy, and electronic warfare, creating a de facto bridge between Silicon Valley innovation and the DoD acquisition pipeline.

Portfolio includes: Anduril, Saronic, Shield AI, SpaceX, Hadrian, and others

Series B and C defense tech rounds are pricing at 10-20x revenue. The primes most likely to acquire the median venture-backed defense company (Lockheed, Northrop, RTX, General Dynamics, L3Harris) have historically paid 2-4x revenue for bolt-on acquisitions. AeroVironment's purchase of BlueHalo for $4.1B priced at roughly 60x revenue, an outlier that may be recalibrating expectations or may simply be an outlier.

Bain's market-sizing work makes the stakes explicit. Their April 2025 estimate put the aggregate valuation of VC-backed defense tech firms (ex-SpaceX, adjusted for dual-use revenue) at $86B; by year-end 2025, that figure had grown to roughly $130B. At a 5x revenue multiple consistent with mature defense tech comps, $130B in valuation implies these companies need $25-30B in combined annual revenue by 2030, about 3% of the roughly $1T in combined US, NATO, and allied procurement and RDT&E spending. Achievable, but it requires every link in the supplier-customer-investor chain to hold.

The NatSec100 is designed to identify where real momentum is building, surface which companies are translating innovation into impact, and illuminate where the system is beginning to work and where it is not.

They captured six seismic shifts that define the operating landscape for the 2026.

1. Washington Rewired the Rules. Executive orders, the Acquisition Transformation Strategy, DRPMs, OSC, EDU, Reconciliation, and the Innovation Adoption Kit.
2. The Battlefield Imperative. Lessons from Ukraine and Iran resulted in increase in AI and autonomy experimentation to central DoW modernization priority and reshaped capital flows, procurement priorities, and operational doctrine.
3. Production as the Bottleneck. Drones, counter-drone systems, munitions, advanced manufacturing, and shipbuilding are all facing the same constraint: production capacity has not kept pace with demand signals or private investment.
4. The Capital Stack is Being Rebuilt. OSC, EDU, equity stakes, DPA-backed financing, leasing arrangements, and large IDIQs represent a new theory of how public and private capital should work together in national security.
5. The Primes Are Adapting. Corporate venture capital funds are expanding significantly. Prime CapEx investment is up roughly 35% for 2026. Compensation ceiling tensions, subcontracting relationships, and M&A activity are all shifting.
6. The Exit Landscape is Maturing. Approximately 20 companies have graduated from the NatSec100 through acquisition or public offering since 2023. These exits represent the clearest evidence yet that defense tech is producing real liquidity events and that the asset class is becoming durable.

The Air Force was such an outlier in 2025 due to 5 major contract obligations:

Sierra Space ($175M) for Tranche 2 Tracking Layer | X-Bow Launch Systems ($129M) for Advanced Integrated Motor Manufacturing | JetZero ($81M) for Blended Wing Body Prototype | DataMinr ($70M) for PADELS | Castelion ($40M) for Low-Cost, Highly Manufacturable Weapon Production

NatSec 100 Top 10

Anduril | Saronic | Sierra Space | Scale AI | Shield AI | Databricks | Axiom Space | Forterra | Castelion | CesiumAstro

Top NatSec 100 Investors (# of investments)

In-Q-Tel 33 | Gaingels 30 | Alumni Ventures 25 | IPO Club 18 | Mana Ventures 16 | Washington Harbor 16 | a16z 15 | Lockheed Martin Ventures 14 | Balerion Space Ventures 13 | Founders Fund 13 | General Catalyst 13 | YCombinator 13

NatSec 100 Company Count by DoW Critical Technology Area

Applied AI 40 | Quantum and BID 27 | Contested Logistics 25 | Hypersonics 4 | Biomanufacturing 3 | Directed Energy 1

Share of Private Capital Raised by DoW Critical Technology Area

Applied AI 75% | Quantum and BID 11% | Contested Logistics 11% | Hypersonics 2% | Biomanufacturing 1% | Directed Energy 1%

Contract Types by NatSec100 Companies

Prime Contracts 43% | OTAs 31% | SBIR/STTR 17% | Subcontracts 9%

They closed with four solid recommendations

1. Measure Adoption, Not Activity
2. Rebuild the Acquisition Workforce as a Warfighting Priority
3. Deploy OSC and EDU Capital into the Hardest Problems
4. Publish What the DoW Actually Plans to Invest In

Our Take: This is outstanding analysis by SVDG and their partners. Our topline takeaway is that the NatSec100 companies still only received a pittance (0.5%) of DoW contract obligations last year.

The possibly more important measure is the signal that DoW has been sending to the investment community thus far - which is that we are not taking advantage of this largesse to field new technologies and capabilities at scale.

We are hopeful there is a bow wave in the next few years with some large programs like CCA and MUSVs getting real production funding and large multi-year munition efforts being approved that is likely to benefit the non-traditional defense community.

As more of these companies and their solutions mature - and get greater visibility as differentiated offerings from what is already fielded - we expect to see increased transitions to scaled production.

This should be enabled by continued acquisition reforms like improved use of OTs, commercial preference, and more competitive approaches.

The big question will be if the volume of orders equals an ROI that satisfies the billions invested to date and if the patience of the capital markets is sufficient to wait for their returns.

it will be interesting to see if the patient investments pay dividends, especially in delivering mission-impactful warfighting capabilities, the ultimate ROI.

Silicon Valley Defense Group (SVDG) 2025 NatSec100 List

• $28.6B in federal awards as prime, subs, OTA, SBIR/STTR - up 18% from 2024
  • $25.8B federal prime
  • $798M federal subcontract
  • $1.8B federal OTA
  • $199M federal SBIR awards
• $70.1B in private capital raised - up 32% from 2024

Source: SVDG NATSEC100 2025

Drivers

• Budget trajectory. The US is tracking toward a >$1T defense budget, with the FY27 request reportedly around $1.5T including reconciliation funding. NATO members are moving toward 5% of GDP defense spending commitments.
• Acquisition reform with teeth, on paper. The SPEED and FoRGED Acts expand multiyear procurement authority beyond legacy platforms into munitions and other systems. The Office of Strategic Capital, now backed by $1.5B in Big Beautiful Bill funding enabling roughly $200B in lending capacity over four years, and the newly launched Economic Defense Unit are both explicitly designed to crowd private capital into production scaling.
• New institutional capital sources. Booz Allen Ventures ($300M fund), Point72's Deterrence Fund ($400M), and JPMorganChase's $1.5T, decade-long Security and Resiliency Initiative (up to $10B earmarked for defense tech, advanced manufacturing, and supply chain resilience) all launched or scaled in 2025.
• Ukraine as proof of concept. PitchBook and SVDG's NatSec100 report both point to Ukraine's drone warfare experience as the operational thesis underwriting the capital rotation toward cheap, attritable, software-defined systems.

Risks

The capital clock doesn't match the budget clock. CNAS frames this as the sharpest structural risk in the sector: venture funds operate on a roughly 10-year capital clock with limited partners expecting liquidity through IPO or M&A. Government demand signals operate on an annual budget cycle that, in 14 of the last 15 years, has been disrupted by continuing resolutions. That mismatch is where hardware-intensive, capital-hungry defense startups are most exposed heading into later funding rounds that require a real program of record to underwrite.

Bain's three-risk framework, echoed in their Breaking Defense op-ed, remains the best short list:

1. Suppliers can't meet the actual capability bar. Contested-environment performance, sustainment, and maintainability in distant theaters are harder than a demo range suggests.
2. Customers don't shift spending or reward risk-taking. Absent real contracting-model changes that reward R&D risk rather than just unit cost, returns won't materialize regardless of technology quality.
3. Investors aren't disciplined or patient enough. Long development timelines and unclear exits mean capital deployed on software-market assumptions will be disappointed by hardware-market realities.

The Reagan Institute's 2026 NSIB Report Card grades the underlying ecosystem a blunt D on “Defense Modernization” (translating innovation into fielded capability at speed) even as it grades “Private Sector Innovator Base” an A-. That gap between innovation supply and fielding execution is, in one letter-grade comparison, the entire thesis of this memo.

For founders:

GTM timing now matters more than it did two years ago. Seed funding is down 58% YoY while late-stage and growth capital is up nearly 90%. Investors underwriting Series B and later want revenue and contract traction, not a roadmap. If your raise depends on a “future program of record” story rather than a signed one, expect a harder diligence process and a longer runway requirement than the 2021-era comps suggest. Dual-use positioning (Groq, Cyera, Form Energy, Helion all show up in Q2's top 10 without being defense-first companies) is a viable path into defense-adjacent capital pools even if defense isn't your primary market.

For investors:

Patience is the scarce resource, not capital. The CNAS capital-clock mismatch is real: a 10-year fund horizon against an annual, CR-prone budget cycle means the exit timeline on hardware-intensive bets is longer and less predictable than software-market intuition suggests. Bain's diligence checklist, a credible end-state business model, a real pricing and sustainment plan, evidence the company can retain culture through scale, is the right filter. The valuation-to-exit-multiple gap (10-20x revenue at entry vs. 2-4x historical prime bolt-on multiples) means underwriting a specific exit thesis, not just a growth story, matters more than it used to.

For ecosystem players (advisors, policy shops, primes):

Procurement reform is the actual lever, not another funding announcement. Every data source in this memo converges on the same conclusion: capital has scaled faster than demand signals. The practical levers worth tracking and advising clients on are multiyear procurement authority (SPEED/FoRGED Acts), Office of Strategic Capital and Economic Defense Unit funding levels, and whether Congress passes appropriations on time. A client's competitive position in this market depends less on which segment they're in and more on whether they can credibly underwrite a path to an actual program of record.