The AI Warplane Is Here. May the Race for Autonomous Air Power Begin.

Boeing’s MQ-28 Ghost Bat is not a replacement for the fighter pilot. It is the first sign of a broader shift in which militaries measure air power by how many autonomous aircraft a single cockpit can command, and Boeing, Anduril, General Atomics, Beijing and a crowded field of allies are all racing to define what that fleet looks like.

At the Farnborough Airshow in July 2026, Boeing did something it has rarely done with a fighter program: it invited a rival’s supplier onto the stage. Program director Glen Ferguson confirmed that Italy’s Leonardo, which had quietly contributed electronic systems to the project for eight years, was now a formal partner on the MQ-28 Ghost Bat, the uncrewed aircraft Boeing built in Australia to fly as a loyal wingman alongside crewed fighters.

The announcement was a small piece of corporate news. It was also a signal of how fast an idea that sounded speculative five years ago, an autonomous jet flying combat missions beside a human pilot, has become the organizing principle of Western air power planning.

Meet the Fighter Jet’s New Best Friend

A loyal wingman, in the terminology now standard across air forces, is a semi-autonomous, jet-powered aircraft designed to operate in formation with a crewed fighter, extending its senses and its weapons load while absorbing the risk that would otherwise fall on a pilot.

The Ghost Bat, at roughly 11.7 meters long with a 7.3 meter wingspan, a range above 2,000 nautical miles and a ceiling above 40,000 feet, is built around that idea: not a stealth fighter without a pilot, but a modular platform that can carry sensors, jamming equipment or missiles depending on the mission. It can be flown from an E-7 airborne early warning aircraft, a ground station or a ship, at effectively unlimited standoff range. In December 2025 it live-fired an AIM-120 AMRAAM guided to its target by data relayed from an E-7, the clearest demonstration yet that an uncrewed aircraft can close the kill chain, not just watch it.

The Future of Air Power Is Human-Machine Teaming

What makes the Ghost Bat strategically significant is less the airframe than the command relationship it implies. Air forces are not building autonomous jets to retire their pilots. They are building them so a single F-35 or future F-47 pilot can, in effect, command a small formation, some of it crewed, most of it not, extending sensor coverage and missile capacity well beyond what one cockpit could otherwise generate.

Boeing has been explicit that the aircraft is designed to complement rather than replace crewed platforms, distributing risk across a formation instead of concentrating it in a single, extremely expensive, irreplaceable pilot. That shift changes what a fighter pilot’s job actually is: less about dogfighting and more about managing a distributed sensor and weapons network in real time, deciding which node in the formation takes a shot, absorbs a risk, or breaks off.

Collaborative combat aircraft are critical to the future of warfare
— Glen Ferguson, MQ-28 program director, Boeing

Training pipelines and command structures have not yet caught up to that reality. Air forces built their promotion tracks, their simulators and their tactics manuals around a pilot flying one aircraft.

The doctrine now being written, in Canberra, Washington and increasingly in Berlin and Tokyo, has to answer a harder question: how much authority does a human in a cockpit retain over aircraft flying thirty miles away at the edge of a jamming envelope, and how is a pilot trained to manage four or five of them rather than one.

Why Militaries Want Cheaper AI Pilots

The economics are blunt. An F-35A carries an average flyaway cost of about $82.5 million, and every hour it flies puts an irreplaceable, years-in-training pilot at risk. The U.S. Air Force’s Collaborative Combat Aircraft program is targeting a unit cost under $30 million, roughly a third of a crewed fighter, for an aircraft that can be sent into the most contested part of the airspace first.

At the cheaper end, attritable designs like the Kratos XQ-58A Valkyrie run closer to three to five million dollars each, aircraft the Pentagon is comfortable losing in a way it would never accept for an F-35. The logic is straightforward: if a nation cannot afford enough crewed fighters to generate mass in a fight with a peer competitor, it buys mass in the form of uncrewed aircraft instead, and attaches them to the crewed fleet it already owns.

The Race to Build the World’s First AI Air Force

Boeing no longer has the field to itself. In June 2026 the U.S. Air Force awarded production contracts for its own Increment 1 aircraft to two rivals: General Atomics for the FQ-42A Dark Merlin and Anduril Industries for the FQ-44A Fury, both selected simultaneously rather than through the usual winner-take-all down-select, preserving competitive pressure on price into full production.

Anduril, a company founded in 2017, has already built a nearly billion-dollar production campus in Ohio capable of turning out 150 aircraft a year. General Atomics brings decades of experience building the Reaper and Predator drone families that quietly did most of the West’s actual combat flying over the past two decades.

Europe is assembling its own field of contenders rather than picking a single design. Airbus has partnered with Kratos to offer a version of the Valkyrie; the German arms maker Rheinmetall has partnered directly with Boeing to offer the Ghost Bat to the Luftwaffe; and the Munich-based defense software firm Helsing is developing its own CA-1 Europa. Boeing’s pitch to Berlin rests on a simple point none of its rivals can yet match: the Ghost Bat is the only allied CCA design that has actually shot something down in a live test.

Who’s Building What • United States: General Atomics FQ-42A Dark Merlin and Anduril FQ-44A Fury, in production for the Air Force’s CCA Increment 1, target cost under $30 million each • Australia: Boeing’s MQ-28 Ghost Bat, the most flight-tested loyal wingman in the world, with Block 2 aircraft due to the Royal Australian Air Force by the end of 2026 • Europe: an Airbus-Kratos offering based on the XQ-58A Valkyrie, Rheinmetall’s partnership with Boeing on Ghost Bat for Germany, and Helsing’s independently developed CA-1 Europa • China: the FH-97 and FH-97A loyal wingman drones alongside the stealthy GJ-11 Sharp Sword and at least two newly revealed tailless uncrewed air superiority designs shown ahead of the 2025 Beijing military parade

Can AI Make Better Split-Second Combat Decisions Than Humans

The case for autonomy is partly a case about speed. In 2023, an Air Force-sponsored simulation reportedly showed an AI system consistently outmaneuvering an experienced F-16 pilot in simulated dogfights, executing turns too fast and precise for a human to counter, and defense officials increasingly argue that in the fastest engagement windows now emerging, a drone swarm closing on an air base, a missile in its terminal phase, there simply is not time for a human to authorize every response. That argument is winning ground inside militaries even as it alarms arms control advocates.

United Nations Secretary-General António Guterres has called fully autonomous weapons “morally repugnant” and pushed for a binding international treaty, a campaign that produced a lopsided but non-binding General Assembly vote in December 2024 in favor of tighter oversight.

Current U.S. and allied policy keeps a human “on the loop” rather than fully out of it, meaning an operator can override a weapon’s decision but is not required to approve every individual engagement, a middle position that satisfies almost no one.

Critics warn of a widening responsibility gap: when an algorithm trained on imperfect data misclassifies a target, it is not obvious who, legally or morally, is answerable. Proponents counter that a human relying on radar and a few seconds of visual identification is hardly a more reliable filter, particularly at the speed modern missile and drone threats now travel.

The AI Arms Race Moves Into the Skies, and Onto Every Other Battlefield

The Ghost Bat is best understood as the aviation branch of a much larger trend: the migration of advances in machine learning and generative AI out of data centers and into physical hardware that fights.

The same autonomy stacks that let a CCA fly formation and identify targets are being adapted for naval drones, ground robots and missile defense systems, and the Pentagon’s push toward a government-owned Autonomy Government Reference Architecture is explicitly designed to let any vendor’s AI software run on any manufacturer’s airframe, ship or vehicle.

That architecture choice matters as much as any single aircraft: it turns autonomy into a software problem that can be upgraded continuously, rather than a hardware problem locked into a single manufacturer’s roadmap for a decade.

What Ghost Bat Means for Future Wars

Boeing has designed the Ghost Bat’s modularity around exactly this flexibility. Its interchangeable nose section can carry different sensor and weapons payloads, letting the same airframe serve as a forward-deployed radar picket, an electronic warfare jammer, a decoy that draws enemy fire away from a crewed jet, or, as the December 2025 AMRAAM test showed, a shooter in its own right.

Collaborative combat aircraft are expected to extend the reach of surveillance and early-warning missions, complicate an adversary’s targeting problem by multiplying the number of things in the sky that might be armed, and let air forces put expendable aircraft into the most dangerous first minutes of a contested airspace fight, the part of the mission historically most likely to get a pilot killed.

Boeing Bets on AI After Years of Commercial Turbulence

For Boeing, the Ghost Bat program lands at a moment when the company badly needs a growth story that has nothing to do with the 737 MAX. Commercial aviation has spent years absorbed by the aftermath of the January 2024 Alaska Airlines door-plug blowout, a crisis that froze production, invited heightened FAA oversight and dragged the company’s free cash flow negative for years. Boeing is only now guiding toward $1 billion to $3 billion in positive free cash flow for 2026, its first sustained cash-positive year since the MAX crisis began compounding in 2019, and even its headline $8.2 billion net income in the fourth quarter of 2025 was driven mostly by a one-time gain from selling its Jeppesen digital aviation unit rather than by underlying operational strength.

Against that backdrop, defense and autonomous systems have become one of the few parts of Boeing’s portfolio where the growth narrative points up rather than sideways. The Ghost Bat is Boeing’s most mature entry in a category the Pentagon has decided it wants at scale, and unlike the fixed-price commercial development contracts that battered Boeing Defense for years, the CCA and loyal wingman market is being built around exactly the kind of iterative, government-partnered, export-friendly model the company needs right now.

A European sale to Germany’s Luftwaffe, brokered through Rheinmetall, would make Ghost Bat Boeing’s first true export win in the category and give the company a foothold in a market it does not currently lead outright, since the Air Force’s own CCA production contracts went to General Atomics and Anduril, not Boeing.

The Export Race for Autonomous Warplanes

Export politics are shaping the program as much as engineering. Japan signed a formal CCA cooperation framework with Australia in April 2026, and South Korea, Singapore and India have all signaled interest, drawn by tight defense budgets and small fleets of expensive fifth-generation fighters that make an affordable, attritable wingman especially attractive.

Britain is exploring integration of European-made weapons like MBDA’s Meteor into any future CCA purchase specifically to reduce its exposure to U.S. arms-export rules known as ITAR, a reminder that even close allies want a mineral-like domestic content, or friendly-country content, threshold before they will commit billions to an American design.

Because the Ghost Bat is built in Australia by an Australian subsidiary of an American company, it occupies unusual middle ground in that debate, marketed as more export-friendly than a purely U.S.-built system while still carrying Boeing’s engineering and supply chain behind it.

China, meanwhile, is not waiting to be invited into that conversation. Its FH-97A drone is, by open-source assessment, closely modeled on the Ghost Bat’s own airframe, and Beijing has already exported earlier-generation military drones to the United Arab Emirates, Pakistan and Serbia, suggesting an eventual export push for its own loyal wingman designs once they mature.

Whoever wins the fight for allied air forces’ procurement budgets over the next five years will not just sell aircraft. They will set the software architecture, the data standards and the training doctrine that a wide swath of the world’s air forces build their future war plans around.

Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our 
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Contact Us to reach a global audience with Impact Newswire.