Welcome back to the #30 edition of The New Defense Post!
In this edition, we’ll cover:
Spotlights: Anduril Secures Up to $20B Enterprise Contract with the U.S. Army; Gecko Robotics Lands a $71M U.S. Navy Deal for AI-Powered Ship Maintenance; Deftak Unveils AI-Guided Drone Munitions That Don’t Need GPS
Fundraising News of the Week: Recent rounds include Smack Technologies’ $32M for military-specific AI models, Chariot Defense’s $34M seed for AI cyber defence of air defence systems, and more.
Bonus Section: Threats in Orbit, on the space defence gap that European founders should be watching.
Spotlights
1. Anduril Secures Up to $20B Enterprise Contract with the U.S. Army
Photo Credit: Anduril
The U.S. Army has awarded Anduril Industries a landmark enterprise contract with a ceiling value of up to $20 billion over ten years. The agreement, announced on March 13, consolidates over 120 previously separate procurement actions into a single contractual framework centred on Anduril’s AI-enabled Lattice software platform, along with integrated hardware, data infrastructure, and technical support services.
The contract includes a five-year base period with a five-year optional extension and is structured around pre-negotiated pricing, which the Army says will reduce administrative overhead and accelerate delivery timelines. The first task order under the new vehicle is an $87 million award to deploy Lattice as the command-and-control backbone for the Joint Interagency Task Force 401, focused on counter-UAS operations. (TechCrunch, Breaking Defense, U.S. Army)
Anduril President Matthew Steckman clarified that the $20 billion figure is a maximum ceiling, not an obligation. Any federal organisation can now order Anduril’s commercially available products through this vehicle, reducing weeks of separate negotiations (Business Insider).
📰 Our Take: This is a structural inflection point. The U.S. military is building an enterprise procurement architecture that treats neoprimes more like platform vendors than traditional contractors.
If Europe’s defense procurement structures followed a similar model, it could dramatically accelerate time-to-field for startups. The Army has now awarded 14 such enterprise contracts in eight months, consolidating 118 separate agreements.
For context, Anduril’s Lattice architecture integrates distributed sensors and autonomous assets into a unified operational layer. Any European startup building open-architecture C2 or sensor-fusion platforms should study this model carefully because Europe might need one.
2. Gecko Robotics Lands $71M U.S. Navy Contract for AI-Powered Ship Maintenance
Photo Credit: Gecko
Gecko’s systems, which include wall-climbing robots, drones, and fixed sensors, build detailed digital twins of each vessel. The Navy says it can identify structural issues up to 50 times faster than manual inspection, and a single flight-deck evaluation has previously eliminated over 3 months of potential maintenance delays. The robots use magnetic adhesion and can get into spaces that are simply inaccessible to human inspectors.
Technical note: Gecko’s AI platform, Cantilever, ingests multimodal sensor data from robots and builds predictive maintenance models. It works across hulls, decks, welds, and components, flagging both current defects and emerging structural issues, including ones invisible to the human eye. Think of it as a living, breathing health model of the entire ship. (Breaking Defense)
🗣 Dave Carter, President, Kratos Defense & Rocket Support Services: “The Princess Anne facility will play a crucial role in supporting our $1.4 billion MACH-TB 2.0 contract… enabling increased test cadence and more affordable flight test opportunities for hypersonic technologies.” (Globe Newswire)
📰 Our Take: Textbook dual-use story. Gecko built its platform in energy and heavy industry, accumulated real operational data, and then crossed over into defence with a proven product.
European founders should take note: build in adjacent civilian sectors first, then enter defence with something that already works. Obviously, don’t do it if you make cruise missiles.
European navies face very similar readiness challenges. Maintenance backlogs, ageing fleets, and a shortage of skilled dockyard workers. The opportunity for predictive-maintenance robotics across European naval programmes is wide open and largely untapped.
The broader point here is that the Navy’s 80% fleet-readiness target by 2027 means the U.S. is actively seeking startups that can compress maintenance timelines by orders of magnitude.
3. Deftak Unveils AI-Guided Drone Munitions That Don’t Need GPS
Image Credit: Deftak
Ukrainian-Estonian startup Deftak publicly presented a new family of AI-guided munitions for UAVs at the Arsenal of Talents exhibition in Kyiv on March 17. The system uses computer vision for active guidance, meaning the munition continuously corrects its trajectory in flight and can strike targets without any GPS dependence.
Each munition packs three onboard components: processing electronics, an optical targeting camera, and a high-explosive warhead. Deftak says the system costs up to ten times less than traditional guided weapons while delivering comparable precision. They have already completed combat testing and are now preparing formal supply contracts with Ukraine’s Ministry of Defence.
Serial production is expected once government approval comes through.
Deftak raised €600,000 from the European Defence Fund Darkstar in 2025. The team includes engineers who previously worked at Google, Facebook, and EPAM. (UNITED24 Media)
📰 Our Take: GPS-denied precision strike is one of the most pressing challenges on the Ukrainian front right now. Electronic warfare degrades GNSS signals over vast areas, making GPS-guided munitions unreliable at best.
Deftak’s approach bypasses this entirely by using onboard optical processing and computer vision for terminal targeting. We should remember that there are more sensors available than just cameras, which could present an interesting challenge for deep tech startups.
For European founders working on guided munitions, drone payloads, or vision-based navigation: track Deftak closely. The fact that the team came from civilian tech rather than traditional defence is also telling.
The transfer of skills from computer vision and ML to guided weapons development is accelerating, and the talent pipeline from Big Tech to defence tech just keeps growing.
Other News
Ukraine launched the Defence AI Centre A1 (UNITED24 Media)
Y Combinator’s latest batch (Winter/Spring 2026) includes a solid crop of defence-focused startups (Y Combinator)
Pilotix is a European drone technology company building end-to-end solutions, from high-performance drones and flight electronics to advanced software systems. With in-house production and high-precision assembly capabilities, Pilotix delivers reliable, scalable platforms for both civilian and defense applications, including surveillance, industrial operations, and specialised missions.
Use code HOBBYDRONEF1 for a discount.
Fundraising News
Amount | Name | Round | Category |
|---|---|---|---|
$550M | Space / Defence Satellites | ||
$71M | Predictive Maintenance Robotics | ||
$32M | Military-Specific AI | ||
$34M | AI Cyber Defence for Air Defence Systems | ||
~€600K | AI-Guided Drone Munitions |
Bonus Section — Threats in Orbit: Europe’s Space Defence Gap
Image Credit: Aldoria Space
Most of the defense tech conversation right now is about drones, counter-UAS, and ground-based autonomy. But 36,000 km above us, a quieter and arguably more consequential contest is playing out.
Russia’s Luch-2 satellite has spent the past two years systematically approaching Western communications platforms in Geostationary orbit, and here is the interesting part: it was SSA startups, not government agencies, that first sounded the alarm. French company Aldoria, US-based Slingshot Aerospace, Switzerland’s S2A Systems, DigitalArsenal.io, and Germany’s Vyoma (developer of the Flamingo SSA constellation) have all played a role in tracking and verifying these movements.
The commercial SSA layer can now detect sophisticated state-operated spacecraft. But detection alone leaves a real gap. Knowing where a threat is tells us very little about what it is actually doing, and nothing at all about how to stop it.
The US already operates the GSSAP constellation for close-range characterisation, and True Anomaly (over $260M raised) has flown prototype inspection spacecraft. Europe’s programmes, IRIS² and EuroQCI, address communications resilience and encryption but leave the problem of physical proximity protection largely untouched.
UK startup Lodestar Space is working to close that gap. Their “bodyguard satellite” concept is built around Mithril, a platform-agnostic autonomy software suite that combines LiDAR, machine vision, and onboard AI to detect, classify, and respond to proximity threats against strategic assets in orbit.
Lodestar’s first orbital validation mission, TESSERACT, is manifested for early 2026 aboard Exotopic’s GMSS mission, aiming to push the sensor suite from TRL 5–6 toward TRL 7. If it works, Lodestar will offer one of the most mature European-adjacent solutions to the proximity-detection problem.
Love these insights? Forward this newsletter to a friend or two. They can subscribe on our website