The 20,000 one-way attack drones are arriving at Fort Benning and forward units, with deliveries ongoing but incomplete. Eleven vendors, Skycutter, Neros, Napatree, ModalAI, Auterion, Ukrainian Defense Drones, Griffon Aerospace, Nokturnal AI, Halo Aeronautics, Ascent Aerosystems, and Farage Precision, are contractually obligated to ship their share of that inventory against approximately $150 million in prototype delivery orders, each vendor hitting a maximum unit cost of $5,000 per drone. This is not a procurement announcement. This is the execution window for the Drone Dominance Program's Gauntlet I phase, which ran from February through July at Fort Benning, where military operators actually flew, assessed, and scored every system before awards were finalized. The Pentagon plans to scale from this initial 20,000-unit run to approximately 340,000 drones across five proven suppliers within the next eighteen months, at a target unit cost of $2,300. That math tells you what the Pentagon cares about right now: not cutting-edge autonomy or sensor fusion, but industrial capacity, supply chain resilience, and the ability to keep a single drone cheap enough to treat as consumable.
The Gauntlet I selection process revealed why. Military operators and Defense Department officials evaluated existing technology against a single criterion: does it work for warfighters today? The DoD program guidance was explicit: "This Gauntlet was about selecting existing technology to meet the needs of warfighters right now, not expecting all drones to do all things." More revealing was the scoring rubric itself. A high-performance drone that required a steep training curve scored lower than a less capable system that soldiers could operationalize faster. One program memo noted: "A high-performance drone that is difficult to operate could score poorly compared to a less capable drone that is easier to use." That preference shaped the winner list. Skycutter, a London-based manufacturer with a track record in volume production, won the Gauntlet. So did Ukrainian Defense Drones, a company building systems tested in real combat for three years. ModalAI and Auterion made the list because their platforms already had established ground-control interfaces and training pipelines. The vendors that did not win were largely the ones pitching next-generation autonomy or sensor packages that required months of pilot training. The signal is unmistakable: the Pentagon is optimizing for deployment speed and operator proficiency, not technology ceiling.
Industrial capacity is now the constraint, and that changes everything about how the competitive landscape works. The Pentagon's two-year plan targets five "proven suppliers" competing on price and reliability, not twenty startups bidding on capability claims. That means the 11 Gauntlet I winners will not all graduate to the full-scale procurement phase. Vendors with demonstrated volume-production experience, supply-chain resilience, and a track record with military customers will advance. Skycutter and Halo Aeronautics already operate factories designed for scale. Ukrainian Defense Drones has active production lines running under combat conditions. ModalAI has the software ecosystem to support rapid training and deployment. The rest face a brutal sorting: prove you can deliver thousands of units per quarter at $2,300 per drone, or watch the contract consolidate with competitors who already can. That is why the July 2026 delivery date matters more than the winning announcement did. Execution reveals which vendors understood the actual requirement, not the fanciest autonomy, but the most reliable supply and the fastest path to warfighter hands.
The implication ripples beyond the Drone Dominance Program itself. Every autonomous-systems company in the defense market is watching how the Pentagon treats the transition from prototype to scale. If Gauntlet I winners can deliver 20,000 units on time with a supply chain intact, the message to investors and program managers is clear: the bottleneck is not technology maturity. It is manufacturing. Companies like Anduril, which built its reputation on autonomous-vehicle autonomy stacks, will have to decide whether to build out production capacity or remain a software provider to primes. Smaller startups that won slots in Gauntlet I but lack factory infrastructure will either raise capital for manufacturing, partner with a contract manufacturer, or exit. The Pentagon, meanwhile, has effectively de-risked the autonomous-systems market for five suppliers: whoever survives the next eighteen months will have a guaranteed path to 340,000-unit procurement and stable margin, which is enough to reshape the whole category. Watch three markers: whether all 11 vendors hit their July delivery targets, whether the training cycle for new operators compresses below two weeks as DoD hopes, and whether the unit cost actually drops to $2,300 by Phase III. If all three happen, the Pentagon will have solved the industrial problem. If any fail, the consolidation accelerates.