The Pentagon is taking longer than ever to put its most expensive weapons into service, with the average time to field a new capability now exceeding 12 years, according to a congressional watchdog report that casts doubt on repeated pledges to speed up delivery.
The US Government Accountability Office, in its 24th annual assessment of major weapons programmes published on 2 July, examined 104 of the Department of War's costliest efforts, a portfolio the department plans to spend more than $2.4 trillion to develop and buy. It found that schedules continued to slip across the board, even for programmes placed on a fast-track route meant to accelerate delivery.
That route, known as the Middle Tier of Acquisition, was designed to prototype or field a system within five years. The GAO found that many programmes entered it before their core technologies were ready, undercutting the purpose of the pathway. Between 2018 and 2025, 18 of the 40 programmes routed through it began with immature technology, and seven of the eight currently on the pathway still require further development.
The department plans to invest at least $49 billion across 23 of its most expensive middle-tier programmes, a portfolio that grew this year as larger and more complex efforts moved onto the route.
"Schedule delays persisted across MDAPs, signaling overly optimistic time frames," the GAO wrote, using its shorthand for major defence acquisition programmes.
The watchdog recommended that the Pentagon require programmes to hold mature technologies before development starts, or to develop immature technologies on a separate track. The department agreed with the recommendation.
Several flagship programmes showed how far timelines have moved. The initial capability of the air-launched Small Diameter Bomb Increment II has slipped by more than a decade against its first full estimate. Boeing Co.'s MQ-25 Stingray, the US Navy's carrier-based refuelling drone, has fallen roughly three years behind, with an operational debut now expected in 2029.
The Space Force was the heaviest user of the fast-track route, accounting for about half of all such spending across the Pentagon. Each of the 13 Space Force programmes examined recorded a schedule setback during 2025.
Among them, the first Next-Generation Overhead Persistent Infrared missile-warning satellite was completed in January, four months later than planned, and now awaits a launch slot no earlier than October because of a crowded manifest. Its sensor payload, built by RTX Corp., ran about $340 million over cost.
The Air Force's Hypersonic Attack Cruise Missile has almost no schedule margin left, programme officials told the GAO, after the service cut its planned flight-test campaign from seven attempts to five. Completing the first three tests is regarded as critical to a decision on whether to begin production in fiscal 2027.
The assessment landed about seven months after Secretary of War Pete Hegseth unveiled the broadest overhaul of Pentagon acquisition in decades. That strategy, released in November, makes speed the organising principle of a rebranded Warfighting Acquisition System and proposes to remove the middle-tier pathway as a distinct process to cut administrative burden.
The GAO has repeatedly questioned whether such changes can deliver on their promise. In reviews published in 2024 and 2025, it found that the Pentagon struggled to carry out earlier reforms intended to accelerate delivery, settling instead into slow and linear development.
Weapons acquisition has featured on the GAO's High Risk List for years, and the office has pressed the department since 2022 to overhaul how it engineers and tests new systems. The department has agreed with many of those recommendations without fully putting them in place.
The delays carry particular weight in the Indo-Pacific, where carrier drones and long-range strike weapons flagged as running late sit among the capabilities the US is counting on to keep pace with China's military build-up.
