The U.S. Air Force plans to invest about $5.8 billion in autonomous drones over the next five years. Its first phase would build about 1,000 to 2,000 aircraft, which could take on different jobs alongside existing fighter pilots.
The plan reflects a push toward what military officials call “affordable mass”: using lower-cost aircraft in larger numbers. The drones could support pilots in flight, gather surveillance, resupply forces, or operate in groups on attack missions.
A lower-cost way to add aircraft
A drone is expected to cost about $3 million, compared with $80 million for an F-35 fighter jet, according to the New York Times. Some drone models would be more expensive, costing up to $25 million, but those aircraft could fly multiple missions and would still cost less than a third of a manned fighter.
That range of prices points to different roles. Cheaper drones could be used for self-destruct missions, while more expensive models could return for repeated flights. The Air Force’s plans include several types rather than one aircraft designed to do everything.
The procurement approach would also differ from that used for fighter jets. The hardware and software for the drones would come from a variety of partners, rather than a single source. That means the broader effort would draw on multiple suppliers as the Air Force pursues its first group of aircraft.
Different drones for different missions
The XQ-58A Valkyrie is one example of the aircraft under consideration. The Air Force plans to deploy drones optimized for surveillance and resupply, as well as wingman aircraft that can accompany fighter pilots on missions.
Some wingman drones could also fly in a swarm to attack. In practical terms, that gives the program several possible uses: an aircraft might provide support to a pilot, help deliver supplies, or take part in a group mission. The source describes these as distinct roles within a wider drone effort.
These functions suggest a shift in how air missions could be organized. Pilots would not necessarily operate alone: a drone could fly with them as a supporting aircraft, while other models would be assigned to tasks such as surveillance. The plan’s emphasis on multiple designs follows from the variety of work the Air Force expects the aircraft to perform.
A flight test for autonomous operations
On July 25, after two years of preparation, the AI-powered XQ-58A Valkyrie successfully completed a test flight in Florida. The test showed that an AI-enabled drone could autonomously solve a “tactically relevant 'challenge problem” while in flight.
That demonstration offers an example of how autonomy could fit into future missions: the aircraft handled a defined problem during flight. The report does not describe the specific challenge, so the result is best understood as evidence of an autonomous capability being tested rather than a full account of how the drone would perform in combat.
Brig. Gen. Scott Cain, commander of the U.S. Air Force Research Laboratory (AFRL), called AI “critical” to future warfare. “AI, Autonomous Operations, and Human-Machine Teaming continue to evolve at an unprecedented pace.” His comments connect the test to a broader focus on pairing human pilots with autonomous systems.
Funding and the next steps
The proposed spending would give the Air Force a large pool of funding for autonomous aircraft, but the budget still faces a step before it can proceed: approval by Congress. Kratos Defense and Security Solutions, the maker of the XQ-58A Valkyrie, plans to compete for the new drone budget if it is approved, according to the New York Times.
The plan’s scale and mix of aircraft make the next phase consequential. The Air Force is seeking drones that can be relatively inexpensive, available in numbers, and suited to different missions. The Valkyrie test demonstrates one piece of that ambition, while the proposed investment would determine how broadly the service can pursue it.