How defense-software, C2, mission-autonomy, and dual-use startups raise capital in 2026 amid Replicator, SDA Tranche awards, CDAO Task Force Lima.
Defense software crossed from contrarian to consensus. Anduril (~$14B), Palantir (public, defense reaccelerating), Shield AI (~$5B), Saronic, Epirus, Chaos Industries, Vannevar Labs, HavocAI, Applied Intuition Defense, Rebellion Defense (challenged), and dozens of SDA/Replicator awardees raised as DoD procurement modernized around SBIR Phase III, OTAs, CSOs, and Commercial Solutions Openings. Investors want a defensible program-of-record path, ITAR/CMMC readiness, and cleared-facility footprint — not another 'AI for defense' deck without a program officer named.
Ukraine, Red Sea, and Taiwan-strait deterrence created durable procurement demand. Replicator scaled from Tranche 1 (all-domain attritable autonomy) into follow-ons. SDA Tranche 3/4 awarded resilient proliferated-LEO military constellations. CDAO Task Force Lima transitioned generative AI into acquisition programs. NATO DIANA and NIF disbursed. FY26 defense authorization emphasized software acquisition reforms. Anduril's Menace, Roadrunner, Fury (CCA); Shield AI's V-BAT and Hivemind; Saronic's Marauder USV; Chaos Industries radar; Epirus HPM all reached scale procurement. Investors treat defense as a category, not a taboo.
Seed: $3-15M for prototype + first SBIR Phase II. Series A: $20-80M for CMMC + program pursuit. Series B: $80-500M+ for hardware/software scale + platform. Reference: Anduril (~$5B+ raised, ~$14B), Shield AI ($1B+ raised, ~$5B), Saronic ($175M+ raised, ~$1B+), Epirus ($350M+ raised), Chaos Industries ($55M A), Vannevar ($90M+ raised), HavocAI (seed/A), Applied Intuition Defense (Applied Intuition ~$6B). Capital-intensive but exit multiples remain strong.
'AI for defense' pitch with no program officer identified. No cleared personnel or cleared-facility plan. Underestimating CMMC/ITAR timeline (12-24 months). Ignoring prime-contractor teaming (Lockheed, RTX, Northrop, L3Harris, Boeing, GD) that gate program-of-record. Weak understanding of DFARS, FAR Part 12 vs Part 15, and Truth in Negotiations. Consumer-tech GTM assumptions applied to a 3-7 year procurement cycle.
Investor directory · Fundraising library · Articles A–Z · Company funding database