Report 208 · Defense Tech
The Drone Dominance Program publishes a leaderboard: ten winners from Gauntlet II, each with a score to one decimal place, splitting orders for 60,000 attack drones. It's the most transparent drone buy I've seen. But electronic warfare was tested in a separate lane and folded into one number with operator ratings and production capability, so the score can't tell you which drone survives jamming.
Most military buying happens behind paper requirements and source-selection files nobody outside will read. The Department of War's Drone Dominance Program (DDP) does something different: it runs vendors' drones through warfighter-designed missions, posts the scores publicly, and places production orders by rank. That deserves credit. It also means a lot of people are now reading a leaderboard as if it were a product review, and as someone who spent his Army career on the jamming side of this fight, I want to be precise about what the number contains.
According to the program's own results announcement (15 September 2026), selection started with a Qualifier at Camp Grayling, Michigan, on 8 to 19 June, where 49 drone companies flew day and night missions. Nineteen advanced. To stay in, each had to fill an order of 120 drones with lethality payloads in about five weeks. Gauntlet II itself ran in August at Fort Carson, Colorado, in two mission areas: Deep Strike, finding and hitting targets from 15 km stand-off, and Close Quarters Battle in confined spaces, with a subset repeated at night.
Then the sentence that matters for my beat. The announcement says: "Additional test lanes assessed controlled kinetic engagement and electronic warfare (EW) performance, with drones flown against counter-UAS systems."
The program lists three inputs. Companies "were scored based on their performance in the event, evaluations by military operators, and demonstrated production and supply chain capabilities." The leaderboard then shows one number per company per mission:
What the public pages do not publish: the weight given to each input, a separate EW score, or the conditions in the EW lane. I looked through the program's results announcement, the leaderboard data, and its other posted announcements. No jammer type, band, power, or distance appears anywhere. That's not a criticism of keeping test parameters quiet; there are good security reasons to. It's a limit on what a reader can conclude.
Coverage hasn't always caught all three inputs. DroneXL's report on the results described the rankings as resting on "mission outcomes and operator ratings" and said the program "has not said whether a vendor's delivery record against an earlier order figures in." The program's own announcement does list production and supply chain capability as a scoring input. How much it counts, and whether it reflects a vendor's delivery record specifically, isn't public.
When it published the first leaderboard in March, the program attached a note from its test directors. It says plainly that the leaderboard "is not a statement about the best drones in the industry, or even the best drones in America," but "a snapshot of how the invited vendors performed against the mission vignettes designed by the warfighters." It also explains that operators got a real vote: "A high-performance drone that is difficult to operate could score poorly compared to a less capable drone that is easier to use."
I think that's the right way to buy drones for soldiers. A drone that's brilliant in a lab and miserable for a tired operator is a bad drone. But it means a 10-point gap on the board might be about ease of use, delivery, or a night run, and the public can't tell which.
Here's the part I can add from the operator side. A drone surviving a jammer is not a property of the drone alone. It depends on the geometry: how far the jammer is from the drone compared with how far the drone's controller is, and how much power each puts out. It depends on whether the jammer covered the bands the drone actually uses for control, video and navigation, and whether the drone could hop, switch links or fly on without satellite navigation. I walked through that geometry in Report 161 on burn-through and in Report 190 on what "jamming-resistant" means.
So a pass in an EW lane is a pass against those jammers, at those ranges, in those bands. Change any of them and the answer can flip. At least one jammer maker that took part, Ukraine's Unwave, has publicized its role: by its account, carried in an article Euromaidan Press labels "Industry Focus," it brought four systems to Fort Carson, alongside 14 drone makers and three EW firms in all, and judged its tasks successfully completed. It said it could not detail the scenarios for security reasons, and the article gives no results for individual drones. That's consistent with what the official pages show: the EW lane existed, and its results live inside the score.
The program's Phase 3 announcement (28 September) moves EW from a side lane to the main stage: both mission sets "will be tested in GNSS- and comms-denied environments." It also sets money and prices: $450 million toward about 60,000 drones, with an option for up to 40,000 more; the Deep Strike price cap rises to $5,000 per drone excluding weaponization (from $4,500), and the Close Quarters Battle price drops to $3,000 (from $3,500). Up to 14 companies will reach Gauntlet III, each after delivering a 400-drone production order in about two months.
If GNSS and comms denial run through every mission, the next scores will say more about jamming than this round's do. They still won't say how much, unless the program publishes a breakdown.
EW lane conditions and results. Not published. Nothing here claims how any named drone performed under jamming.
Scoring weights. The three inputs are official; their weights are not public.
Unwave's account. Vendor claims, carried in an article the outlet labels "Industry Focus." Not independently verified, and used only to show that dedicated jamming systems took part.
The Drone Dominance leaderboard is a genuine step toward buying weapons on demonstrated performance, and its test directors are honest about its limits. Read it for what it is: one number per company that blends mission results, operator judgment and production capability, with electronic warfare measured in a separate lane and not broken out. "Top of the leaderboard" doesn't mean "best under jamming." Phase 3's GNSS- and comms-denied missions should change that, and the useful thing to ask for next is a published EW sub-score.
U.S. Army combat veteran (Counter-IED / Electronic Warfare), peer-reviewed researcher in microwave spectroscopy, and founder & CEO of Shroombiosis. Consults on laboratory operations, AI, and supplement formulation.