Replicator Year Two: What's Actually in the Water

The Pentagon promised thousands of attritable autonomous systems in 18 to 24 months. Two years on, the fleet is real but uneven — and the honest scorecard is more interesting than either the boosters or the skeptics admit.

The Promise, Restated

When Deputy Secretary of Defense Kathleen Hicks unveiled Replicator in August 2023, the commitment was specific: field attritable autonomous systems at scale of multiple thousands, in multiple domains, within 18 to 24 months. The initiative was never about buying new hardware. It was a process bet — proof that the Department of Defense could move commercially-derived unmanned systems from contract to combat-relevant inventory faster than the traditional acquisition cycle, which typically runs a decade or more for a program of record.

The budget was modest by Pentagon standards: roughly $500 million per year, about $1 billion across FY2024 and FY2025, pulled largely from reprogramming and service commitments rather than new appropriations. That figure matters. Replicator was a forcing function, not a procurement tsunami.

What Actually Fielded

Tranche 1, disclosed in stages through 2024 and 2025, leaned on a short list of systems the services already knew:

- AeroVironment's Switchblade 600 loitering munition, backed by a contract ceiling of nearly $1 billion awarded in August 2024 — the single largest Replicator-adjacent award and the clearest demand signal from Army customers. - Anduril's Ghost-X rotorcraft, an intelligence and surveillance platform already in operational testing. - PDW's C100 quadcopter, a rucksackable strike and ISR drone that earned selection after competitive fly-offs with infantry units. - Maritime surface systems, confirmed but largely classified in detail, with reporting pointing to small USVs from vendors including Saronic Technologies. - Coyote interceptors and associated counter-UAS kits, which pulled double duty as both Replicator systems and fielded air defense.

By late 2025, the department publicly claimed it had met the thousands-fielded threshold. Independent verification was never really possible — the count blends delivered hardware, training articles, and systems purchased by the services on Replicator's timeline but outside its direct funding. The honest grade: the initiative hit its number by its own accounting, and it genuinely compressed timelines, but "thousands" includes a lot of short-endurance quadcopters rather than thousands of deep-strike attritable airframes.

What Slipped

The gaps are more instructive than the deliveries.

Attritable one-way attack airframes with meaningful range — the systems most relevant to the INDOPACOM "hellscape" concept articulated by Admiral Samuel Paparo — arrived slower and in smaller numbers than the loitering munitions. Cheap, long-range, mass-producible strike drones turned out to be a manufacturing problem, not a design problem, and the U.S. industrial base entered 2026 still scaling toward rather than sustaining the rates that Ukrainian production lines, producing on the order of millions of FPV-class airframes annually by Ukrainian government claims, treat as routine.

Maritime numbers also lagged the rhetoric. Small USVs are in the water, but the fleet measured in hulls is in the dozens-to-low-hundreds range across programs, not the swarm imagery of the early briefings. Adversary-shaping quantities remain a 2027-and-beyond proposition.

Tranche 2 and the Counter-Drone Pivot

Replicator's second tranche, announced in late 2024, marked a deliberate pivot toward counter-small-UAS for base defense — an acknowledgment that the most immediate unmanned threat was inbound, not outbound. Selections reported across 2025 drew from vendors including Anduril, Fortem Technologies, and established air-defense primes, layered onto systems like the Coyote interceptor and the Army's FAAD C2 architecture that stitches sensors and effectors together.

The logic was economic as much as operational. The exchange ratios from the Red Sea and Ukraine made the case brutally clear: a $100,000-class interceptor engaging a $20,000-to-$50,000 Shahed-type drone is sustainable; a $2 million missile doing the same job is not. Tranche 2 institutionalized that math inside the budget process, and by 2027 the services are treating C-UAS layering — guns, interceptors, electronic warfare, high-power microwave systems like Epirus' Leonidas — as standard base-infrastructure procurement rather than an urgent operational need.

What INDOPACOM Actually Asked For

The demand signal from the Pacific has been consistent and specific: range, survivability in a contested electromagnetic environment, and sustainment without contractor logistics tails. Admiral Paparo's hellscape framing — saturating the Taiwan Strait with thousands of unmanned systems to impose delay and cost in the first weeks of a conflict — survived contact with acquisition reality better than most concepts, but with a quieter addendum the services now write into requirements: the systems must be produced at rates that assume high attrition and no resupply for weeks.

That addendum is where Replicator's real legacy is being written. The initiative's sunset at the end of Hicks's tenure was expected; instead, its mechanisms — the Defense Innovation Unit as an accelerator, senior-leader reviews that broke logjams in weeks rather than years, and the willingness to buy before requirements were "perfect" — have been absorbed into the services' normal processes. Quietly, each service has drafted its own after-action language. The common threads: commercial-first selection works, production capacity must be contracted before it is needed, and operator feedback loops measured in weeks beat test reports measured in quarters.

The scorecard that matters is not how many drones the Pentagon fielded by a date, but whether it learned to buy the next batch faster than the adversary learns to kill the current one.

What to Watch Through 2027

- Production-rate contracts: whether Switchblade 600 and follow-on loitering munitions move from hundreds to thousands of units annually, and whether unit prices fall toward the $50,000-and-below band the services say they need. - Tranche 3 signals: reporting points to maritime strike and undersea systems; watch for formal selection announcements and, more important, delivery schedules. - USV hull counts: whether Pacific Fleet exercises in 2027 involve dozens of surface vessels operating as coordinated groups or continue at demonstration scale. - C-UAS exchange ratios: interceptor unit costs against the Geran-class threat remain the governing number; watch Coyote follow-on pricing and directed-energy cost-per-shot disclosures. - Institutionalization: if Replicator-style rapid-fielding reviews appear in the FY2028 budget request as permanent line items, the experiment becomes doctrine. If they fade, it was a two-year anomaly — an expensive, instructive one, but an anomaly nonetheless.