A real Air Force general said the quiet part out loud about AI and human pilots
In July 2026, DARPA and the U.S. Air Force flew the first standard, front-line F-16 modified to fly under full AI control, part of the VENOM Autonomy Kit program, building directly on 2023 and 2024 tests where an AI agent flew a real F-16 in an actual within-visual-range dogfight against a human pilot. This isn't a simulation or a concept video. It's a real aircraft, in real airspace, with a real DARPA program manager confirming the milestone. And in a public panel, the Air Force's former head of acquisition put the stakes in blunt terms most people haven't heard: if a human being is in the loop, in a future fight against AI, you will lose.
// Field Guide
Get The Human Control Gap: $8
What actually stops an autonomous weapon from making a kill decision, and what doesn't. The real policy versus the common myth.
Get the guide: $8 instant download →DARPA's Air Combat Evolution program first proved the concept using the X-62A VISTA, a heavily modified, one-of-a-kind experimental F-16, flying real dogfights against a human-piloted jet in 2023 and 2024. VENOM is a deliberate next step: taking the same AI control approach and installing it on standard, unmodified front-line F-16s, without altering the jet's core software, so a pilot can switch between human and AI control mid-flight. DARPA's own program manager, Brig. Gen. James Valpiani, described it as building the infrastructure needed to develop trusted, autonomous air combat capabilities at scale, not just prove a single demonstration works.
Then-Air Force acquisition chief Andrew Hunter, formerly Frank Kendall, addressed this directly and publicly: the critical problem on the battlefield is time, and AI will be able to do much more complicated things much more accurately and much faster than human beings can. His conclusion, stated plainly at a public defense forum, was that a human physically in the control loop during a future engagement against AI-piloted aircraft will lose, and that human supervision, watching over what the AI does without directly intervening in real time, is the only version of oversight that remains viable.
This establishes that the U.S. military is not merely researching autonomous air combat as a distant concept, it is actively flying real, AI-controlled fighter jets on operational-type airframes right now, and that senior defense officials have publicly and explicitly stated that direct human control will be a battlefield disadvantage in future engagements. This is confirmed by DARPA's own program announcements and on-the-record public statements from Air Force leadership.
What this does not establish is that any AI-controlled aircraft is currently authorized to independently select or engage a target without a human decision in the loop. VENOM and ACE remain explicitly framed as human-on-the-loop testing programs; the debate DARPA's own officials are having in public is about what happens next, not a claim that fully autonomous lethal authority already exists today.
Program: DARPA/U.S. Air Force VENOM Autonomy Kit, first flight July 2026. Built on DARPA's Air Combat Evolution program, first autonomous dogfight demonstrated April 2024.
You just read one document. Pro shows you every document that contradicts it.
The Contradiction Engine cross-references any two records in the archive and shows you where the government’s own story changed. The Evidence Chain View maps how one program became the next, sourced at every step. Citations generated correctly, every time. Your own notes, saved against the document.
Start your free month →This site costs real money to run. If it’s useful to you, you can help keep it going →