What if the most important part of a sixth-generation fighter is not the fighter at all? That question now sits at the center of modern combat aviation. For years, sixth-generation airpower was described through familiar hardware language: more stealth, more range, new engines, exotic sensors, and perhaps directed-energy weapons. Those ideas still matter, and many of them were visible in earlier U.S. thinking about a future fighter, which emphasized deeper sensor fusion, a heavier command-and-control burden, and even the prospect of a “digital wingman.” But the center of gravity has shifted. The leading edge is no longer just an aircraft design. It is the ability to connect a crewed platform to autonomous partners, update tactics through software, and create combat mass without buying only more exquisite manned jets. That shift is no longer theoretical.
The Air Force is now flight-testing mission autonomy on its first Collaborative Combat Aircraft prototypes, pairing the government-owned Autonomy Government Reference Architecture with General Atomics’ YFQ-42A and Anduril’s YFQ-44A. The architecture matters as much as the airframes because it is designed to prevent vendor lock and let algorithms move across platforms. In plain terms, the service is trying to separate the drone’s autonomy software from its airframe. That is a major change from the traditional fighter model, where hardware and software often remain tightly bound to a single prime contractor’s ecosystem. Col. Timothy Helfrich described the goal directly: “We are instead building a competitive ecosystem where the best algorithms can be deployed rapidly to the warfighter on any A-GRA compliant platform, regardless of the vendor providing the algorithm.” That is the real sixth-generation challenge: adaptation speed
Both autonomy vendors now publicly tied to the effort, Collins Aerospace and Shield AI, are being judged on how their software handles combat-relevant behaviors, mission updates, and coordination with crewed aircraft. The Air Force has also confirmed it plans to decide on an Increment 1 winner by the end of 2026, with the selection covering both the aircraft and the mission autonomy software. That framing says a great deal about where air combat is headed. A future fighter pilot is becoming less of a lone tactical operator and more of a mission commander, assigning tasks, managing sensor coverage, and using uncrewed aircraft as forward scouts, weapons carriers, or decoys.
Older debates about whether a sixth-generation platform would be manned, unmanned, or optionally manned now look incomplete. The emerging answer is a blended force in which the human stays in the system, but the job changes. Earlier Air Force thinking stressed that a human brain still offered superior real-time judgment and situational awareness in fast, ambiguous combat. Nothing in the current CCA push overturns that. Instead, it reinforces it by moving routine flying, positioning, and some tactical execution onto software while preserving human control over mission intent and weapons employment parameters.
The scale of the ambition is striking. The service has discussed at least 1,000 CCAs in the broader program, while current prototypes are already feeding operational experimentation and logistics planning rather than remaining isolated test articles. Weapons integration flights have begun, autonomy packages are flying, and the drones are being treated as part of the larger NGAD family rather than as side projects. In that environment, the crewed sixth-generation aircraft becomes the central coordinator of a distributed formation, not simply the most capable individual platform in the force.
That may be the clearest sign that the fighter’s new job has arrived. The race is still about air dominance, but increasingly it will be won by the force that canupdate software and tactics faster than hardware can be redesigned.
