An AI-controlled fighter jet, the X-62A VISTA, goes up against a human-piloted F-16 jet during a test at California’s Edwards Air Force Base.
It’s unclear if the X-62A outperformed the human-controlled fighter jet, or if that was even a goal for the dogfight. But DARPA and the US Air Force are hailing the test as a major breakthrough because it showed that AI-controlled jet fights are possible. Specifically, the AI-controlled jet violated no training rules during the test flights, suggesting it can operate safely.
During the test, the X-62A did have human pilots onboard who could monitor and take over when necessary. “However, test pilots did not have to activate the safety switch at any point during the dogfights over Edwards,” said DARPA program manager Lt. Col. Ryan Hefron. “We have to be able to trust these algorithms to use them in a real-world setting.”
The big question is whether the US Air Force plans on deploying AI-controlled fighter jets in real military operations. For now, DARPA has only said it’ll continue test flights with X-62A this year. Still, Air Force Secretary Frank Kendall is hailing the tests as a “transformational moment.”
“The potential for autonomous air-to-air combat has been imaginable for decades, but the reality has remained a distant dream up until now. In 2023, the X-62A broke one of the most significant barriers in combat aviation,” he said in a statement.
The X-62A is part of DARPA’s Air Combat Evolution (ACE) program, which notes a human-piloted fighter jet could exact more damage by “orchestrating multiple autonomous unmanned platforms from within a manned aircraft.”
“In particular, ACE aims to deliver a capability that enables a pilot to attend to a broader, more global air command mission while their aircraft and teamed unmanned systems are engaged in individual tactics,” DARPA added.