AI Pilots X-62A in First Aerial Dogfight

The U.S. Air Force Test Pilot School and the Defense Advanced Research Projects Agency worked together to test breakthrough executions in artificial intelligence algorithms using the X-62A VISTA aircraft as part of DARPA’s Air Combat Evolution (ACE) program. (Image: Air Force Test Center)

The U.S. Air Force Test Pilot School and the Defense Advanced Research Projects Agency have reported another major breakthrough for the artificial intelligence (AI) algorithms or "agents" that have been flying their modified F-16 test aircraft known as X-62A in a series of flight tests since 2022.

In an April 17 press release, the Air Force Research Laboratory (AFRL) reported on the use of AI algorithms to complete their first ever demonstration of an AI-controlled fighter jet against a human-controlled fighter jet for a simulated in-air dogfight. The AI-controlled dogfighting breakthrough occurred at Edwards Air Force Base, California in September 2023, however, AFRL and DARPA have not released details about it until now.

VISTA was developed by Lockheed Martin Skunk Works in collaboration with Calspan Corporation for the USAF TPS. Built on open systems architecture, VISTA is equipped with software that allows it to mimic the performance characteristics of other aircraft.

The Air Force summarizes “dogfighting” as aerial combat occurring between fighter jets at close range. The demonstration is considered a major breakthrough for the use of AI algorithms to control a fighter jet not only in a dog fighting scenario, but it proves that the algorithms can complete nearly every type of operation or maneuver that a human pilot can.

William Gray, Chief Test Pilot, Air Force Test Pilot School, and other engineers conduct software updates to the X-62 Variable Stability In-Flight Simulator Test Aircraft at Edwards Air Force Base, California, Aug. 3, 2022. (Image: U.S. Air Force, Giancarlo Casem)

“It's very easy to look at the X-62A ACE program and see it as under autonomous control, it can dogfight, but that misses the point," said Bill Gray, the Test Pilot School's Chief Test Pilot. "Dogfighting was the problem to solve so we could start testing autonomous artificial intelligence systems in the air. Every lesson we're learning applies to every task you could give to an autonomous system."

Secretary of the Air Force Frank Kendall will soon take flight in the X-62A to "personally witness AI in a simulated combat environment during a forthcoming test flight at Edwards," according to AFRL.

The dogfighting AI algorithm demonstration came shortly after a previous breakthrough for the algorithms, when AFRL and DARPA reported on the X-62A's December 2022 execution of 12 flight tests of maneuvers such as one-on-one beyond-visual-range, or BVR and engagements against a simulated adversary.

"In less than a calendar year the teams went from the initial installation of live AI agents into the X-62A’s systems, to demonstrating the first AI versus human within-visual-range engagements, otherwise known as a dogfight. In total, the team made over 100,000 lines of flight-critical software changes across 21 test flights," AFRL notes in its release.

During the tests, the X-62A is always flown with safety pilots onboard that have the ability to independently disengage the aircraft's flight management computer from the control of the AI agents. During the dogfights, the safety pilots did not have to deactivate the AI agents' control of the aircraft. The tests featured the AI agents in "nose-to-nose engagements where the dogfighting aircraft got as close as 2,000 feet at 1,200 miles per hour," according to AFRL.

Col. James Valpiani, Commandant of the Test Pilot School, described the AI-piloted dogfights as a "paradigm shift" for the use of AI algorithms to control fighter jets — and possibly other aircraft in the future.

"We've fundamentally changed the conversation by showing this can be executed safely and responsibly,” Valpiani said.