Deep Signal: @shieldaitech: The U.S. Air Force just made autonomy software as important as the aircraft. The @usairforce selec
U.S. Air Force selects Shield AI's Hivemind autonomy software for Anduril's Fury collaborative combat aircraft, decoupling software from airframe procurement.
- 1,000+ Potential CCA fleet requirement over decade-long procurement horizon USAF cost targets $25–30M per unit
- Q2 2026 Fury production line start at Arsenal-1 facility, Pickaway County, Ohio
- 5 Vendor CCA field competing for USAF downselect
- HQ
- Costa Mesa, California, United States
- Founded
- 2017
- Employees
- 1,000
- Funding Total
- $6.3B
- Products
- Fury·Lattice·ALTIUS-700M
- Competitors
- Boeing·Lockheed Martin·Northrop Grumman·General Atomics
Shield AI’s Hivemind Wins CCA Software Slot — What the Air Force Just Told the Market
Product Portfolio — Anduril
Signal Activity — Anduril
Deal History — Anduril
Competitive Positioning — Anduril
What Happened
The U.S. Air Force has selected Shield AI’s Hivemind autonomy software to run on Anduril’s Fury (YFQ-44A) collaborative combat aircraft for the Collaborative Combat Aircraft (CCA) flight demonstration phase. This is a software-on-hardware pairing between two private defense technology companies — neither of which is a traditional prime — competing within a five-vendor USAF CCA field that also includes Boeing, Lockheed Martin, and Northrop Grumman.
The CCA program targets procurement of autonomous wingman aircraft to fly alongside crewed fighters, with the Air Force seeking hundreds of units at a unit cost target reportedly below $25–30 million. Shield AI, founded in 2015 and having raised approximately $500 million in total funding, has positioned Hivemind as a pilot-replacement autonomy stack capable of operating in GPS-denied, communications-degraded environments. Fury is currently at prototype deployment status, with Arsenal-1 production in Pickaway County, Ohio targeted for Q2 2026.
Why It Matters
This selection establishes a structural precedent: the Air Force is explicitly decoupling autonomy software from airframe procurement. That is a significant architectural decision. Historically, DoD aircraft programs bundled avionics, flight control, and mission software with the platform vendor. By selecting Hivemind as a separable software layer running on Fury, the Air Force signals it may treat autonomy stacks as interchangeable or competitively sourced components — similar to how commercial cloud infrastructure is procured independently of application software.
This pairing strengthens both companies’ CCA downselect positioning. Anduril gains a validated autonomy software partner with demonstrated GPS-denied flight credentials; Shield AI gains a Group 5 airframe to demonstrate Hivemind at operationally relevant scale.
Market Implications
The Hivemind-Fury pairing creates a template for future autonomous platform competitions. If replicated across other CCA vendors or follow-on programs, this model could fragment the traditional prime contractor monopoly on flight-critical software, opening procurement pathways for specialized autonomy vendors. For RTX, Lockheed Martin, and Boeing, the precedent suggests autonomous software may become a competitive vulnerability if internal development lags specialized players like Shield AI.
The Q2 2026 Arsenal-1 production timeline is aggressive. Demonstration of manufacturing scale and software maturity at that pace will be closely watched by both the Air Force and competing CCA teams.