About a year after it first showed the concept, Shield AI has built the real aircraft. The company says X-BAT 001 is finished. It is the first full-scale prototype of the company's jet-powered autonomous strike aircraft, which takes off and lands vertically by sitting on its tail. The airframe has left the company's factory in Frisco, Texas, and is on its way to a flight test site in Newton, Kansas. Shield AI says it will fly there before the end of 2026.
The announcement coincided with the IQPC International Fighter Conference in Munich, where Shield AI's director of CCA Europe described strong European interest in the program. The rollout is also the first time the public has seen how the full-scale design compares with the half-scale model shown in April.
From announcement to airframe in about a year
Shield AI announced X-BAT in October 2025. In a Nov. 4 blog post, quoted by Air & Space Forces Magazine, Armor Harris, the company's senior vice president of aircraft engineering, said the first full-scale prototype was complete in Frisco and on its way to Newton. That puts the time from public reveal to finished airframe at about one year. The company presents that speed as the main argument for the program.
"We cannot take decades to develop an aircraft when the need is already here," Harris said.
Andrew Hawkes, Shield AI's director of CCA Europe, gave Janes a similar timeline on Nov. 4. He said the concept went public 12 months earlier and that the first prototype left the Frisco factory the week before. Hawkes described X-BAT 001 as a full-size aircraft with full autonomy and full integration of its General Electric F110-GE-129 engine, including thrust vectoring. He said it would fly at Shield AI's Kansas test range by the end of the year.
Shield AI calls the Frisco plant X-Forge 1 and the Newton flight test site X-Forge 2, according to TWZ. TWZ also reports that X-BAT 001 is the first of two full-scale prototypes the company plans to build.
What changed since Sea-Air-Space
TWZ compared photos of the finished prototype with the half-scale model Shield AI showed at Sea-Air-Space 2026 in April and with earlier renderings. It found the overall configuration broadly similar, with a few notable differences.
The planform. The April model had already moved away from the original "cranked kite" wing shape to an arrowhead with a straight leading edge. TWZ says X-BAT 001 is broadly reflective of that model.
The nozzle. Previously shown models and renders had the thrust-vectoring nozzle inset in the rear fuselage. On X-BAT 001 it extends significantly to the rear, beyond the trailing edges of the wings. TWZ reports that the nozzle provides pitch and yaw control only, and that Shield AI has said bleed air routed through thrusters in the wings will provide roll control.
New dorsal features. The prototype has two pronounced bulges toward the rear of the central body, one on either side of the dorsal spine, which do not appear on past models or renderings. It also has two circular openings on top of the spine covered with mesh screens. Their purpose is unknown, and Shield AI declined to comment when TWZ asked. TWZ says they could be tied to the drone's propulsion arrangement. A Shield AI spokesperson told TWZ that X-BAT 001 is a development aircraft, so the design will evolve as the company tests and refines it.
The Navy's role
The program has US Navy involvement. According to Air & Space Forces Magazine, the Navy made subject matter experts available to Shield AI's team, and they collaborated on technical risks including wind tunnel, engine and thermal testing.
TWZ describes the Navy's interest as a significant vote of confidence and says the Navy hopes to see the drone become a real operational capability on a very aggressive timeline. Sharing technical expertise is not the same as buying aircraft, and neither source reports a funding commitment from the Navy for X-BAT.
The test plan
Shield AI's path to first flight is staged, as tail-sitter programs usually are. According to Air & Space Forces Magazine and TWZ, the first step has the prototype horizontal on its launch and recovery vehicle, running the engine, before it is raised to a vertical position. Next, the prototype is suspended from a crane for testing, so it does not crash if the engine stalls or another problem emerges. TWZ quotes Harris saying a very large crane is being set up at the Kansas test site. Untethered flights come last.
Hawkes's promise of flight by year-end leaves about eight weeks after rollout. Janes does not say whether "fly" means a tethered hover or an untethered flight. That will matter when the company announces the event.
The performance claims
Shield AI's claims for X-BAT are ambitious. Air & Space Forces Magazine reports that the company says the aircraft will be stealthy, will fly at supersonic speed (unlike other autonomous drones competing in Collaborative Combat Aircraft programs), and will have a combat radius of 1,000 nautical miles. Because it takes off and lands vertically, the company says it could operate from ships, islands and austere sites.
None of this has been shown in flight yet. TWZ notes that the core X-BAT concept still needs to be proven, and that there have been several unsuccessful attempts to operationalize broadly similar designs in the past, which Shield AI has acknowledged. The prototype is meant to start testing whether the combination works.
European interest
The announcement coincided with the Munich conference. Hawkes told Janes that European interest has been building since the late-2025 reveal. Shield AI has a dedicated CCA Europe director, which suggests it sees the continent as a real market and not only a place to exhibit. Janes did not name specific countries or programs.
Why It Matters
Many uncrewed combat aircraft now in development need runways. Runways are fixed and easy to target, and many of the places where these aircraft might be needed do not have them. X-BAT is betting that a jet able to launch from a ship deck, an island or a rough clearing is worth the large engineering cost of vertical flight. With X-BAT 001 built, that bet now has hardware behind it.
The prototype also tests a claim about how quickly aircraft can be developed. Shield AI says it went from public concept to full-scale airframe in about a year, and it is promising flight within weeks of rollout. If it meets that schedule, its argument that combat aircraft do not need decade-long programs gets stronger. If it slips, the timeline will draw closer scrutiny. Either way, the Kansas tests will put the first real data behind claims that so far exist only on paper.
