In August 2023, Deputy Secretary of Defense Kathleen Hicks announced Replicator: "multiple thousands" of all-domain attritable autonomous systems, fielded within 18-24 months.
It's 2026. The Congressional Research Service says we got "hundreds."
Not thousands. Hundreds.
What Actually Got Built
To be fair, Replicator accomplished something real. AeroVironment's Switchblade-600 loitering munitions entered production. Anduril's Altius-600, Ghost-X, and Dive-LD got contracts. Performance Drone Works' C-100 was selected. Seven software companies were brought in for autonomy and command-and-control.
Approximately $1 billion was committed across FY2024-2025. The Defense Innovation Unit proved that non-traditional acquisition pathways can work. Some of these systems are genuinely capable and operationally relevant.
But "hundreds" vs. "thousands" matters. Because the entire point was speed. The thesis was: we can move as fast as startups. We can buy autonomous systems like Ukraine buys drones — in volume, fast.
And we moved faster than normal. Just not fast enough.
Meanwhile, Ukraine
While we were fielding hundreds, Ukraine's drone industrial base went from 7 manufacturers before the invasion to over 500 today. They produce approximately 4 million drones per year. Fifteen companies build fiber-optic FPV drones that are immune to electronic warfare jamming. The cost asymmetry this creates is devastating — a $500 FPV drone vs. a $3.73 million interceptor.
Saronic went from founding to a $392M Navy contract with delivery in under 12 months. Anduril delivered Ghost Shark prototypes to Australia on time, on budget, in three years — while Boeing's ORCA sat at $885M and counting.
The benchmark isn't "faster than the Pentagon's historical average." The benchmark is "fast enough to matter." And by that measure, hundreds when you promised thousands is not fast enough.
If your doctor says you need to lose 50 pounds and you lose 12, nobody calls that a win. The defense acquisition system's version of success is still too slow for the threat.
Replicator Lives On — Sort Of
The initiative didn't die. It was reportedly reorganized as the Defense Autonomous Warfare Group (DAWG) and placed under SOCOM. Admiral Paparo at INDOPACOM confirmed in December 2025: "It's alive." The DAWG reportedly focuses on larger, longer-range attack drones for Pacific operations.
Replicator 2.0 — focused on counter-drone systems — was consolidated under Joint Interagency Task Force 401 in August 2025. Its first actual purchase came on January 11, 2026: two Fortem DroneHunter F700 systems, scheduled for delivery by April 2026.
Two systems. In January 2026. Three and a half years after the original "thousands by August 2025" announcement.
The Honest Assessment
The problem isn't that Replicator failed. It's that "hundreds" was treated as success when the benchmark was "thousands."
We're incrementally faster at a time when we need to be exponentially faster. The acquisition reforms worked on the margins. But the margins aren't where the threat lives.
Secretary Hegseth's "Drone Dominance" initiative, announced in July 2025, runs in parallel — focused on smaller FPV-style drones for small units. That's promising. But the lesson of Replicator 1.0 should be clear: set the goal, measure against the goal, and be honest when you miss.
The most dangerous thing in defense acquisition isn't failure. It's calling failure success and moving on.
Was Replicator a win, a failure, or something in between? I think the answer depends on what you compare it to — our own historical baseline, or the pace of the actual threat.
Quick Answers
Did Replicator fail?
The article argues that Replicator improved speed at the margins, but failed relative to its own “thousands” benchmark. It was progress, not mission accomplishment.
Why does this matter beyond one Pentagon initiative?
Because Replicator is a case study in whether the defense acquisition system can field autonomous mass quickly enough to stay relevant against adversaries who iterate on much shorter timelines.
Sources: CRS IF12611 (Sept 2025), DIU.mil/replicator, Breaking Defense (Dec 2025), DefenseScoop (Feb 2026), Washington Times (Nov 2025), defense.gov JIATF-401 announcement (Aug 2025)