Air Force To Build First CCA Drones!

The U.S. Air Force just picked the first “robot wingmen” it plans to buy at scale, and the real fight now is over who gets to own their brains.

Story Snapshot

The Air Force chose General Atomics and Anduril to build the first production Collaborative Combat Aircraft fighter “wingman” drones.
Separate firms will compete to supply the autonomy software, so the brain and the body of each drone can be swapped like phone apps.
The program aims for at least 150 drones this decade on the way to a fleet in the thousands, reshaping air combat and budgets.
The same openness that fights vendor lock-in could also spread risk, blame, and technical failure across many private players.

America’s first AI wingmen move from PowerPoint to production

The Air Force has now crossed a line that defense programs rarely return from: it is putting real money on contract to build operational Collaborative Combat Aircraft, not just prototypes for air shows.^[2] General Atomics and Anduril each received engineering, manufacturing, and production deals for Increment 1 airframes, covering the first three lots and at least 150 drones by decade’s end.^[2] For taxpayers, that means CCA is no longer a science project; it is a program that will demand sustainment dollars for years.

These drones are not tiny quadcopters; they are jet-powered “loyal wingmen” meant to fly with F-35s and future sixth-generation fighters in serious combat.^[4]^[7] Air Force leaders talk about “affordable mass”—fielding many attritable aircraft rather than a small handful of gold-plated jets.^[4] General Atomics brings decades of experience with Reaper-style unmanned aircraft, while Anduril’s Fury design, born from its Blue Force Technologies purchase, represents a fast, high-performance fighter-like drone.^[6]^[14] That pairing alone signals a shift: old-guard industry sharing the stage with Silicon Valley-style defense startups.

Why the software competition matters more than the metal

Buried inside the announcement is the truly radical piece: the Air Force is treating the autonomy software as a separate competition from the airframe.^[1]^[2]^[5] Anduril, Shield AI, and Collins Aerospace will spend two six-month phases battling to supply the “brain” that flies these aircraft and helps them fight.^[2] That decision fits a larger Modular Open Systems Approach, which aims to decouple hardware from software so the government can change vendors without scrapping entire fleets.^[5]

Evidence that this is more than buzzwords already exists. Anduril’s YFQ-44A has flown a sortie where it completed mission tasks with Shield AI’s Hivemind autonomy, then switched mid-flight to Anduril’s own Lattice system and repeated the same objectives before landing safely.^[11] That single flight showed in public what the Air Force has wanted for years: a combat aircraft that can swap autonomy stacks instead of being locked to one proprietary brain. From a conservative, competition-minded view, that is the right instinct—keep vendors honest, keep options open, and avoid captive monopolies.

From concept art to designated fighter prototypes

Just a few years ago, CCA lived mostly in animations and speeches. That phase is over. The service formally designated General Atomics’ prototype as YFQ-42A and Anduril’s as YFQ-44A, giving them the same “F” fighter mission prefix used on crewed jets, but with a “Q” to mark them unmanned.^[7] Within about 18 months of key awards, General Atomics flew YFQ-42A in August 2025 and Anduril flew YFQ-44A in October 2025, both now in live test campaigns.^[3]^[5]

Flight testing has already moved beyond simple patterns in the sky. The Air Force is integrating inert air-to-air missiles on Anduril’s jet and preparing live weapons shots.^[4]^[5]^[15] Service officials say the goal is to have at least one CCA design operational by about 2030, with a possible fleet of roughly 1,000 aircraft or more over time.^[1]^[2]^[4]^[12] That is the kind of scale that changes how many pilots you need, how you plan missions, and how you think about losing aircraft in a shooting war.

Promise and risk: affordable mass, complex accountability

Supporters argue that attritable autonomous drones are the only way to counter large adversaries without bankrupting the country.^[4]^[23]^[24] Losing a $25 million drone hurts less than losing a pilot and a $100 million fighter, and a large swarm of semi-autonomous systems can saturate enemy defenses.^[6]^[24] Market forecasts back this trend: analysis projects the military drone market more than tripling between 2026 and 2031 as forces pivot toward unmanned systems and swarm tactics.^[24]^[23] That is a rare case where strategic logic and budget reality point in the same direction.

The U.S. Air Force awarded production contracts to General Atomics and Anduril to build its first fleet ‌of semi-autonomous Collaborative Combat Aircraft (CCA), moving a program that began just over two years ago from prototype to full-scale manufacturing.

USAF awarded…

— Aeronews (@AeronewsGlobal) June 18, 2026

Yet the same modular, contractor-heavy model that fights vendor lock-in also complicates responsibility. Legal scholars already warn that mixing increasingly autonomous weapons with heavy outsourcing spreads decision-making across many actors and blurs who is accountable when something goes wrong.^[22] The CCA ecosystem will involve prime airframe contractors, software vendors, cloud providers, and data analysts—many in the private sector. From a common-sense conservative lens, that demands tight oversight: strong rules of engagement, clear human-on-the-loop control, and real technical testing before these drones carry live weapons on America’s behalf.

Why this moment matters more than the headline suggests

Some skeptics point out that the Pentagon often showers prototypes with attention, only to cut them when budgets tighten.^[27] That caution is healthy. But CCA is already past the usual “paper airplane” stage—Congress is being asked for nearly a billion dollars in early procurement and advance buys, and the Air Force is planning multiple lots, not a one-off demo.^[6]^[2] Combined with Ukraine’s lesson that cheap drones can shape a war’s outcome, this suggests unmanned combat aircraft are not a fad but the next baseline.^[21]^[26]

For citizens who care about national defense and fiscal restraint, the key questions now are simple. Will modular autonomy and open architectures actually keep costs and vendors in check, or will complexity itself become the new form of lock-in? And will leaders insist on real human judgment at the point of lethal decision, or slide into a world where opaque algorithms decide who lives and dies? The General Atomics–Anduril win is not the end of that debate. It is where it truly begins.