WASHINGTON—The United States is losing high-value, multirole unmanned drones at an alarming rate. Reporting on August 13 suggested that the US military has lost upward of one-fourth of its MQ-9 Reaper drones in conflicts in the Middle East so far this year. In the first forty days of Operation Epic Fury, a reported twenty-four MQ-9 Reaper drones were lost. Six months into the conflict, the tally is forty-five. US MQ-9 losses during Operation Rough Rider against the Houthis in 2025 were also substantial. At a congressional hearing in May, Air Force Deputy Chief of Staff for Plans and Programs Lieutenant General David Tabor told Congress that the current Air Force MQ-9 fleet is now composed of 135 aircraft—down from 206 in 2024.
However, the real cost will be on the battlefield. MQ-9s are in high demand, with Air Force Chief of Staff General Kenneth Wilsbach calling them “the most valuable player” in Operation Epic Fury. Historically, MQ-9s have been the platform of choice for strikes on high-level targets, including Islamic Revolutionary Guard Corps (IRGC) Qods Force leader Qasem Soleimani in January 2020 and al-Qaeda leader Ayman al-Zawahiri in July 2022. While the Air Force plans to eventually phase out the Reaper, the tempo of current operations could stress US readiness with depleted MQ-9 inventories.
MQ-9s can perform several functions. These medium-altitude unmanned aerial vehicles can provide round-the-clock surveillance of enemies, conduct precision strikes and electronic warfare, and assist with combat search and rescue. Perhaps most importantly, they can allow military decision-makers to carry out missions without putting US service members directly at risk.
The United States uses Reapers in “orbits,” cycling aircraft in and out of an area to maintain continuous coverage. The Forward Defense Initiative’s Iran War Tracker documented that the United States has consistently sustained ten MQ-9 orbits throughout Operation Epic Fury. It is impossible to estimate the exact number of MQ-9s over Iran given the distances involved and how aircraft were grouped for missions. However, the significant MQ-9 losses, high operational tempo, and the Air Force chief of staff’s comments indicate that MQ-9s have maintained a large and constant presence over the theater.
The US Air Force leadership’s response to the shortage is twofold: buy existing systems in the immediate term and ask industry to create less expensive alternatives as part of a broader modernization effort. Specifically, a recent US Senate hearing revealed that Air Force leadership wants the next generation of multirole drones to be modular—that is, built around a less expensive, mass-produced skeleton whose modules can be tailored to the mission. As part of that effort, the Air Force recently partnered with the Defense Innovation Unit (DIU) to scope out possible successors.
In an era of rapidly evolving drone warfare, it is worth asking whether such a drone would solve the military’s unfavorable cost–exchange problem. Even as it seeks out a cheaper successor, the Air Force’s current emphasis on another multirole platform suggests that the service is reluctant to depart from a proven legacy system. But against more competent adversaries who can contest the airspace, the multirole model may simply lack the resilience required to keep drone equipment losses at an acceptable level.
A modular system still concentrates multiple capabilities in individual drones, leaving a single point of failure. The airframe may be less expensive, but the specialized equipment it carries could still make each drone costly to lose. Nor does modularity provide much redundancy—backup capacity if one drone goes down—unless several drones can perform the same mission.
Instead, the Air Force should break the multirole model into its component missions. Unmanned Aerial Vehicle (UAV) technology now exists to perform many of the MQ-9’s functions in more affordable, more scalable ways. For example, several AI-enabled autonomous drones could work collaboratively to scan an inaccessible area for a missing service member. A similar network could track an enemy target, then relay its location to loitering high-altitude drones that can execute a strike.
Such a disaggregated model could fit into existing US Air Force operations. Especially with advances in artificial intelligence (AI), drone connectivity and command-and-control requirements could still be handled by personnel who previously operated MQ-9s.
Disaggregation could also flip the cost-exchange problem onto the adversary. A crowded airspace filled with many less expensive drones presents a new targeting problem for militaries with limited stocks of surface-to-air missiles and and counter-drone capabilities. Countries like Iran can mass-produce simple, cruise missile–like drones such as Shaheds. But can they produce enough interceptors to shoot down large numbers of smaller loitering drones?
Specialized drones could also make the broader system easier to test and upgrade. Unlike a modular platform, changes to one specialized drone would not necessarily force tradeoffs across unrelated capabilities. If a cheaper or more effective solution is found for a specific point in the mission, that component can simply be replaced independently. This reduces testing requirements, which can be costly, and allows the system to adapt at higher speeds.
The Air Force and DIU should therefore focus their search for the next generation of drones on how the system can be enhanced—not simply the individual platform.
Ultimately, the multirole model is, and will be, an extremely useful system—just not necessarily against more sophisticated adversaries. The MQ-9 was the brainchild of a US defense industry reconciling with the small wars of the 1990s and 2000s. Those wars were fought against significantly less organized and less technologically competent enemies than today’s US adversaries. The high-low mix favored the United States because its enemies could never compete at the high end.