What has been a giant leap toward rewriting the future of warfare, the US Air Force just completed successful testing of an AI-controlled fighter jet pitted against a human pilot in real-world dogfighting. This would have effectively shown a quantum leap into applying artificial intelligence in flying, a field ruled by human skill and intuition until now.
Run as part of the DARPA ACE program, the test pitted the X-62A VISTA, an F-16 modified to fly the test, against a human-piloted F-16. The X-62A, controlled by machine learning agents, did “high-aspect nose-to-nose engagements” with a human-piloted F-16, closing to within 2,000 feet at 1,200 miles per hour.
The Department of Defense called the effort “the first AI vs human within-visual-range engagement,” but it remained mum on whether the AI-controlled aircraft came out on top. Still, the successful flight test in and of itself is considered a huge success.
“Beginning in December of 2022, that was the first application of machine learning agents to control the flight path of fighter aircraft,” said Col. James Valpiani, commandant of the Air Force Test Pilot School.
Developed and refined across 21 unique test flights, the system underwent a run from December 2022 through September 2023. During that period, over 100,000 lines of flight-critical software changes were made to finalize the tools.
While it wasn’t about flexing the combat capability of AI, it has shown that AI, at such complex levels of warfighting, can do a host of activities safely and ethically.
“In advance of formal verification methods for AI-based autonomy, the team pioneered new methods to train and test AI agent compliance with safety requirements, including flight envelope protection and aerial/ground collision avoidance, as well as with ethical requirements including combat training rules, weapons engagement zones, and clear avenues of fire,” the DOD program update said.
The biggest supporter of US military adoption, Secretary of the Air Force Frank Kendall, described the capability as “transformational.” He said he also plans to fly in an F-16 later this year in autonomous flight mode.
Further informing the decision to push forward with the Air Force’s collaborative combat aircraft program was the progress of the ACE program toward the development and fielding of next-generation autonomous drones for a variety of missions, most notably related to counter-air operations.
ACE is solely conducted activity by a few academic and industry partners including Calspan, Cubic Corporation, EpiSci, Lockheed Martin Skunk Works, physics, Shield AI, the University of Iowa Operator Performance Laboratory, Johns Hopkins Applied Physics Laboratory, the MIT Computer Science and Artificial Intelligence Laboratory, and the MIT Lincoln Laboratory.
Though impressive in its own right, an AI-controlled fighter jet is only an early example. As AI continues to expand in its capabilities, so too are its possible uses in military aviation likely to grow, and arguably could one day change the nature of warfare.