3/570 → 19/779; +1.91 percentage points of formation share.
What stands out across this batch is how many companies are converging on the exact same problem from different angles: counter-drone defense. Arlo Industries, AICE, 9 Mothers, Maquoketa Research, Surtr Defense Systems, Earendil Robotics, IMPACT Drones, and Greypoint Industries are all building sensing, fusion, or interception layers aimed at the same class of aerial and underwater threats, which suggests the market is less about who has a drone and more about who can see, classify, and stop someone else's. That crowding helps explain why the closely related "Drone Defense" formation is growing even faster, at 6.17× versus this formation's 4.63×, effectively pulling volume out of the broader Defense category into a narrower specialty. Meanwhile a second, smaller thread is quietly working on the industrial base itself, with GUILD and Isengard Industries focused on manufacturing capacity and sovereign production rather than any single weapon system, and Tenet Industries pushing design-for-manufacturing into attack drones, a bet that the winners will be decided by who can build fastest and cheapest rather than who has the cleverest sensor. The live work on autonomy testing and counter-autonomy attribution points to a real gap underneath all of this: as systems like 9 Mothers' EDDA and Surtr's ParallaxOS get fielded, nobody yet has a reliable way to tell malfunction from spoofing from deliberate adversarial control in the field. Much of this cluster only became buildable after cheap small-drone warfare in Ukraine turned counter-UAS from a niche military program into an urgent, funded procurement category starting around 2022, while falling costs of autonomy stacks and edge compute since then made it possible for small teams, not just primes, to compete for that money.
Autonomous systems need repeatable evaluation across contested sensing, communications loss, adversarial behavior, mission changes, and human command boundaries.
Map program test requirements against scenario generation, hardware-in-the-loop evaluation, mission rehearsal, red-teaming, and evidence delivery.
As autonomous systems proliferate, operators may need to distinguish malfunction, spoofing, adversarial control, and intentional behavior quickly enough to respond safely.
Research operational demand for sensor-fusion provenance, behavioral attribution, countermeasure verification, and post-event evidence across autonomous domains.
With a dozen hardware-heavy defense startups (Icarus, Arlo Industries, Surtr Defense Systems, Tenet Industries) emerging in just two batches, none appear to be certification or compliance-focused; these companies will need to prove airworthiness, ITAR compliance, and milspec conformance to sell to primes and government, creating demand for a shared testing/certification layer rather than each building it in-house.
Check back around June 2027 for new YC batches, to see whether independent evidence for "Independent testing, certification, and compliance infrastructure for defense hardware startups" has appeared.
Companies like Null Labs, Icarus, and AICE are building physical aerospace/industrial hardware, but as defense-tagged 'Industrials' startups multiply (9 Mothers, Maquoketa Research, DAIVIN!), the bottleneck shifts from design to rapid, secure, domestic small-batch manufacturing that avoids foreign supply chains — a role not yet visible among the listed companies.
Check back around August 2027 for new YC batches, to see whether independent evidence for "Defense-specific manufacturing and supply chain shops for small-batch hardware" has appeared.
Seeing Systems (AI, Hardware, Drones, Aerospace) and IMPACT Drones (AI, Drones, Swarm Robotics) both pair AI with physical sensing platforms, suggesting a growing need for standardized perception/fusion software that can be reused across the many hardware-first entrants (Earendil Robotics, Icarus) rather than each building bespoke vision stacks.
Check back around May 2027 for new YC batches, to see whether independent evidence for "Sensor fusion / perception software layer for drone and robotics fleets" has appeared.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Included in YC's Defense membership during the measured recent window.
Measured-window members of Defense that could not be assigned to a narrower shared YC tag or industry without forcing a classification.
A primary internal company segment within Defense, derived deterministically from the shared yc tag “Drones”. This is a classification of current window members, not a causal layer or a separate time trend.
A primary internal company segment within Defense, derived deterministically from the shared yc tag “Aerospace”. This is a classification of current window members, not a causal layer or a separate time trend.
These share a word with “Defense” but are tracked as separate categories. The real overlap below is the actual share of Defense’s companies also found in that category — not a guess.
Cambrian Signal, “Defense,” publication yc-2026-10-02-semantic-v8, snapshot October 2, 2026.
cambrian.jafar.com/formation/defense