Shield AI Wants to Build the X-BAT Fighter that Doesn't Need a Runway

By Garth Calitz

Autonomous combat aircraft moves closer to first flight, promising to take the runway out of the military aviation equation. Whether the engineering can deliver on the ambition is another question entirely.

The military aviation world has never been particularly short of ambitious ideas. Flying wings, stealth bombers, hypersonic missiles and aircraft capable of taking off vertically have all moved from drawing boards into operational service. Now American defence technology company Shield AI wants to take things a step further with an autonomous combat aircraft that does not need a runway, an aircraft carrier catapult or even a pilot.

Its X-BAT programme is moving towards its first flight, following the completion of a full-scale prototype. The aircraft represents an ambitious attempt to combine vertical take-off and landing (VTOL), jet-powered performance, autonomous flight and substantial weapons-carrying capability in a single platform.

If Shield AI succeeds, it could offer military commanders an entirely different approach to deploying combat air power. If it doesn't, it will at least provide a rather expensive demonstration of why conventional runways have remained popular for more than a century.

Shield AI first revealed the X-BAT concept in October 2025. In just over a year, the company has progressed from an ambitious design proposal to a full-scale prototype, designated X-BAT 001. The aircraft was unveiled at the International Fighter Conference in Munich in November 2026 before being transported from Shield AI's development facility in Frisco, Texas, to its flight-test centre in Newton, Kansas. Two additional prototypes are expected to follow as the development programme gathers momentum. The company is targeting an initial flight before the end of 2026.

The X-BAT is a substantial aircraft, weighing approximately 22,000 lb (9,979 kg). Its proposed performance figures are equally ambitious.

Specification Published figures

Wingspan 11.9 m

Maximum range Over 2,000 nm

Combat radius 1,000 nm

Maximum payload 2,722 kg

Service ceiling Above 50,000 ft

Powerplant GE F110-GE-129E

Crew None

These are manufacturer-projected capabilities rather than independently verified flight-test results.

The X-BAT's most distinctive feature is its tail-sitting configuration. Rather than using a conventional runway, the aircraft will be positioned vertically on a specially designed mobile launch-and-recovery system. Its General Electric F110-GE-129E turbofan, equipped with an axisymmetric thrust-vectoring exhaust nozzle, will provide the thrust needed to lift the aircraft vertically.

The F110 is no lightweight experiment in propulsion technology. It belongs to the same engine family powering the F-16 and F-15, although the X-BAT installation incorporates modifications specifically developed for its unusual operating environment. The aircraft transitions from vertical flight into conventional forward flight, where it is intended to operate at supersonic speeds. Landing, however, is where things become particularly interesting.

The X-BAT must execute a high-angle-of-attack manoeuvre, described as similar to the famous Cobra manoeuvre, before returning to a vertical attitude. Additional roll controls and wingtip jets will help stabilise the aircraft as it positions itself for landing.

This is not something one would want to attempt with a poorly calibrated flight-control computer. More importantly, the engineering challenges are substantial. Sustained vertical operation requires modifications to the engine lubrication system, while the heat generated by the exhaust presents obvious difficulties when operating from ships.

Shield AI has already worked with GE Aerospace on engine modifications and conducted integrated testing of the thrust-vectoring nozzle. However, successful engine testing on a test stand and repeated vertical take-offs and landings in operational conditions are two very different achievements.

At the heart of the X-BAT is Shield AI's Hivemind autonomy software. The system is designed to allow the aircraft to operate without a human pilot onboard and with limited or no continuous communication with ground controllers.

This is particularly important in modern combat environments where GPS signals can be jammed, communications disrupted and traditional command networks compromised. Hivemind is intended to allow the X-BAT to navigate, assess its operating environment and execute assigned missions autonomously.

The aircraft is being developed for multiple roles, including air-to-air combat, precision air-to-ground strikes and electronic warfare. It will carry weapons internally as well as on external hardpoints. Shield AI also envisages multiple aircraft operating collaboratively under human supervision, effectively allowing one operator to control a formation of autonomous combat aircraft.

That introduces an entirely new dimension to air combat. Instead of risking a highly trained pilot and an expensive crewed fighter in every engagement, commanders could deploy autonomous aircraft to undertake particularly dangerous missions. The technology raises obvious questions about human control over lethal decisions, accountability and the rules governing autonomous weapons. These are questions that defence manufacturers and governments will need to address rather than simply leaving them to the software engineers.

The X-BAT is attracting serious attention from the US Navy, which is supporting its development through the Runway Independent Maritime Expeditionary Strike (RIMES) programme. The programme has received approximately US$200 million in US government funding, matched by a further US$200 million from Shield AI. The Navy's interest is understandable. Conventional fixed-wing naval aviation depends heavily on aircraft carriers, which are extraordinarily expensive platforms requiring enormous logistical support.

An aircraft capable of operating from smaller vessels, expeditionary sea bases and remote locations could provide additional strike capability without requiring a conventional carrier. Shield AI is also marketing the aircraft internationally, with reported interest from European and Asian defence establishments.

An aircraft that could operate from dispersed locations would offer a different approach to survivability. However, the X-BAT is not a simple solution to the challenges facing smaller air forces. Its engine, sophisticated electronics, weapons systems and specialised launch-and-recovery equipment would demand substantial financial and technical support.

For the South African Air Force, the more immediate lesson may be the growing importance of unmanned systems, autonomous mission planning and the ability to operate effectively in environments where traditional communications and navigation infrastructure cannot be taken for granted.

The days when unmanned aircraft were regarded primarily as surveillance platforms are rapidly disappearing. The X-BAT is undoubtedly one of the more intriguing military aircraft projects currently under development. It combines several technologies that have individually been demonstrated, but bringing them together into a practical, operational combat aircraft is another matter.

Vertical take-off is not new. The Harrier and F-35B have already demonstrated that jet-powered aircraft can operate without conventional runways. Autonomous flight is also well established, while thrust-vectoring technology has been around for decades.

• 

• 

What Shield AI is attempting is to combine all these capabilities in an aircraft that can operate autonomously, fly at supersonic speeds, carry a meaningful weapons payload and recover vertically without the support of a conventional airbase.

There are also some obvious questions. How reliable will the vertical recovery system be in turbulent conditions? How will the aircraft cope with repeated shipboard operations? What will its maintenance requirements look like? And perhaps most importantly, what will it cost to acquire and operate?

The company is working towards a mission-equipped aircraft in 2028, with production targeted for 2029. Those are ambitious timelines for a platform that has yet to demonstrate an untethered flight.

The X-BAT remains a promising prototype rather than a proven combat aircraft. But the underlying concept deserves attention. Modern air forces are increasingly concerned about the vulnerability of large, expensive airbases and the concentration of combat aircraft in relatively few locations.

If Shield AI can demonstrate that a supersonic autonomous aircraft can reliably take off and land vertically from a small mobile platform, it could fundamentally change how future air power is deployed. The runway may not be disappearing from military aviation, but it could soon find itself competing with a rather unconventional alternative. And somewhere, an airfield operations manager is probably wondering whether this means he can finally stop worrying about runway maintenance. For the rest of us, the first flight should be considerably more interesting than the paperwork.