- The U.S. Navy believes that AI will allow it to shift to mostly uncrewed aircraft.
- The shift would make carrier air wings cheaper and offload dangerous missions away from human crews.
- In order for that to happen, the Navy believes it must have “trust in autonomy”—and be confident that AI will do what it is told.
The U.S. Navy believes that drones are the future on the carrier flight deck, but first it has to trust them.
The service, which has predicted that up to 60 percent of the carrier air wing of the future will consist of drones, believes it must come to trust autonomous systems in the same way it can trust people—to do exactly as ordered and carry out missions. The service has a number of drones both in the pipeline and under development that could revolutionize the carrier air wing.
A Foundation of Trust
Rear Admiral Stephen Tedford, program executive officer for Unmanned Aviation and Strike Weapons, recently told attendees at the Sea-Air-Space 2023 conference that the Navy must come to trust AI-powered weapons systems to do their jobs. “If we have trust in autonomy,” Telford explained, “we can then make the move to truly artificial intelligence and in the future of the air wing.”
Right now, the Navy can trust its human pilots to perform missions. If a carrier sends a squadron of strike fighters into defended airspace to perform a mission, the commanders of the mission don’t question the dedication of the pilots to do their utmost. The mission may fail, but it won’t be for a lack of trying.
Unmanned aircraft powered by artificial intelligence introduce uncertainty into the situation. Will the AI’s developers foresee every reasonable possibility? Will a software bug lead to an AI canceling a mission, returning home without even trying? Will an AI, faced with a dilemma and operating autonomously, make the correct decision? These are all questions the Navy has never had to face before, but if it can solve them, it will reap huge huge benefits.
A Fix For the Carrier’s Woes?
The Navy has big plans for unmanned aircraft, and it might just solve the problems of naval aviation. Crewed navy strike fighters, particularly the F-35C Lightning II, have steadily become more unaffordable over time. The F-35C costs $90 million each, and the two squadrons of F-35Cs assigned to each carrier cost a total of $2.16 billion. That’s the cost of just 24 out of 70 or so aircraft that make up a carrier air wing, and a Ford-class carrier itself costs $12.7 billion.
Drones promise to save the carrier air wing from itself. A drone doesn’t require the life support systems, human interface, and other features pilots require, cutting both weight and cost. Drones typically cost dramatically less than manned aircraft: the Kratos XQ-58A Valkyrie drone costs just $7.5 million each. While the Valkyrie is less effective than a F-35C, teaming the two together could allow the Navy to cheaply augment each manned jet with several unmanned jets.
The Navy first flew an unmanned aircraft from a carrier in May 2013, when an X-47B unmanned drone was launched and recovered from the USS Bush. Unfortunately for drone enthusiasts, the X-47B never entered service and the Navy took a step back. A new drone, the MQ-25 Stingray, will join the fleet in 2026 as an aerial refueling drone. The MQ-25 will extend the reach of manned aircraft—a leap forward, but not exactly the fighting drone enthusiasts had in mind.
Last July, USNI News reported that the Navy had three more drones in development: a spy drone capable of operating in defended enemy airspace, a strike drone capable of penetrating enemy airspace, and a smaller multi-mission drone that could carry extra missiles for fighter jets, an electronic warfare package, or a command and control package for other drones.
Crawl, Walk, Run
The MQ-25 Stingray’s mission as an aerial refueling drone may have been a disappointment but it was part of a long game. A refueling plane has a fairly limited mission set, and the Navy can learn to start trusting AI on non-combat aircraft. Once it has reasonable confidence that the MQ-25’s software is sound, it can proceed to platforms with more dangerous missions, that pilots might have to trust in extreme circumstances.
About five MQ-25s will serve with each aircraft carrier. The Navy’s own Next Generation Air Dominance (NGAD) program, separate from the Air Force’s NGAD program, will consist of both manned and unmanned aircraft. The average carrier has 40 to 44 manned strike fighters. If each NGAD fighter is accompanied by an unmanned wingman, we could eventually see up to 12 uncrewed fighters, plus five MQ-25s, and the reconnaissance, strike, and multi-mission drones also in development. Within 20 years, up to half of the planes operating from a carrier flight deck could be unmanned.
The Takeaway
While AI will probably never completely replace human pilots on the carrier flight deck, it will probably evolve within our lifetimes to the point where it could take on all the missions of a carrier air wing. The U.S. Navy has been slow to embrace unmanned aircraft, but the skids are greased to ensure that, once it happens, unmanned planes could take over the air wing very quickly. The Navy has very wisely decided that trust is the key factor here, and it must come to trust unmanned planes to do their jobs as effectively as manned planes.
Kyle Mizokami is a writer on defense and security issues and has been at Popular Mechanics since 2015. If it involves explosions or projectiles, he's generally in favor of it. Kyle’s articles have appeared at The Daily Beast, U.S. Naval Institute News, The Diplomat, Foreign Policy, Combat Aircraft Monthly, VICE News, and others. He lives in San Francisco.
