TLDR: In a landmark event for autonomous aerial combat, an AI-controlled X-62A VISTA, a modified F-16, engaged a human-piloted F-16 in a series of real dogfights over Edwards Air Force Base. The tests, part of DARPA’s Air Combat Evolution (ACE) program, demonstrated the AI’s ability to perform complex maneuvers at supersonic speeds, with its performance reportedly on par with its human counterpart. This breakthrough signals a significant step towards future AI-driven military aviation.
A groundbreaking series of aerial dogfights took place above Edwards Air Force Base in Kern County, California, marking a pivotal moment in the development of autonomous combat technology. An AI-controlled X-62A Variable In-flight Simulator Aircraft (VISTA), a heavily modified F-16, went head-to-head against a human-piloted F-16 in what the U.S. military described as a groundbreaking test. While the news item provided was dated September 13, 2025, reports indicate these historic flights and their results were primarily announced in April 2024, following tests conducted in December 2022 and culminating in dogfights in late 2023.
The tests were a core component of the Defense Advanced Research Projects Agency’s (DARPA) Air Combat Evolution (ACE) program, initiated in 2019. The program aims to develop AI-driven solutions for challenges in aerial combat, with a long-term vision for collaborative combat aircraft—unmanned, AI-controlled jets operating as autonomous wingmen alongside piloted fighters.
During the engagements, the AI-controlled F-16 demonstrated remarkable capabilities, reaching supersonic speeds and executing aggressive maneuvers, including vertical ascents and nose-to-nose passes. The aircraft reportedly came within 2,000 feet of its human-piloted counterpart, a testament to the precision and responsiveness of the AI system. According to the Air Force, the AI’s performance was “about even” with the human pilot’s, showcasing its proficiency in dynamic combat scenarios.
Lt. Col. Ryan Hefron, the ACE program manager for DARPA, emphasized the safety aspect of the trials, stating, “The purpose of the test was to demonstrate we can safely test these AI agents in a safety-critical air combat environment.” Human pilots were present in the X-62A cockpit to intervene if necessary, but no interventions were required, and no safety rule violations were recorded during the flights. This successful demonstration underscores the growing trust in AI for critical flight systems.
Former Air Force Secretary Frank Kendall highlighted the strategic implications, noting in 2024 that the U.S. aims to deploy weapon-equipped uncrewed aircraft by 2030. He even expressed his intent to ride in an autonomously flown F-16, demonstrating confidence in the technology’s readiness. Bill Gray, chief test pilot at the USAF Test Pilot School, noted that the ACE program is about more than just aerial combat, hinting at broader applications for AI in aviation.
Also Read:
- Iran’s National AI Regulatory Body Poised for Official Approval
- Microsoft Commits $80 Billion to AI Development, Intensifying Rivalry with OpenAI and Google
The AI system behind the X-62A operates on principles similar to conventional machine learning, analyzing vast datasets of historical flight scenarios to determine optimal real-time solutions. The program progressed from extensive flight simulator scenarios, where AI agents developed by Heron Systems achieved a 5-0 whitewash against an experienced Air Force F-16 pilot, to real-world flight tests. This progression from simulation to live combat scenarios without human intervention marks a significant milestone in military AI and the future of air warfare.
