Shield AI’s Hivemind Wins the CCA Software Layer — on Anduril’s Fury Airframe

The Air Force’s decision to pair Shield AI’s Hivemind autonomy software with Anduril’s Fury (YFQ-44A) for Collaborative Combat Aircraft flight demonstrations signals that the DoD is deliberately disaggregating the autonomy stack from the airframe — and that no single company will own the full CCA value chain.

This is structurally significant for both firms. Anduril, which has $6.3B in total funding and commenced YFQ-44A production at its Arsenal-1 facility in March 2026, built Fury around its own Lattice autonomy platform. The selection of Shield AI’s Hivemind — a competing autonomy stack — for CCA flight demonstrations suggests the Air Force is either running a competitive software bake-off, hedging against single-vendor lock-in, or both. Shield AI, which closed a $2B raise (Series G: $1.5B plus $500M preferred equity) at a $12.7B valuation in March 2026, has positioned Hivemind as hardware-agnostic; its prior deployments on F-16s and MQ-25 testbeds gave it credibility the Air Force could not ignore. The Fury-Hivemind pairing is a flight demonstration, not a program of record — but demonstrations at this stage of CCA development directly inform downselect decisions.

Dimension	Anduril (Fury / Lattice)	Shield AI (Hivemind)
CCA Role	Airframe manufacturer (YFQ-44A)	Autonomy software provider
Autonomy Platform	Lattice (fielded across USSF, Army, USAF)	Hivemind (F-16, MQ-25 demonstrated)
Recent Valuation	~$14B verified (Series F, Aug 2024)	$12.7B (Series G, Mar 2026)
Recent Funding	$1.5B Series F + reported $2.5B follow-on	$2B (Series G + preferred)
Production Status	YFQ-44A production started Mar 2026 at Arsenal-1	Software only; no airframe manufacturing
Key Risk	CCA downselect; Arsenal-1 ramp execution	Airframe dependency; no hardware moat

For Anduril, the development is a double-edged signal. The Fury airframe being selected as the demonstration platform validates the hardware — Arsenal-1 production is real, and the Army’s $87M JIATF-401 Lattice contract (first task order under a $20B vehicle) confirms Lattice’s C2 role in counter-drone missions. But Hivemind’s insertion into the CCA software layer means Lattice does not have a guaranteed position on the Air Force’s most strategically important autonomous aircraft program. Anduril’s wide-moat thesis rests partly on Lattice creating cross-domain integration stickiness; a Fury airframe running a competitor’s autonomy software in Air Force demonstrations is a direct test of that thesis. The outcome of the CCA downselect — expected to narrow five vendors to fewer — will determine whether Anduril captures both the hardware and software revenue streams or only the former.

BOTTOM LINE

Procurement officers and program analysts tracking CCA should treat the Fury-Hivemind pairing as evidence that the Air Force intends to maintain competitive pressure on autonomy software independent of airframe selection — and should monitor whether Lattice or Hivemind achieves preferred status before the program-of-record downselect.

Confidence: MODERATE — The demonstration pairing is confirmed by both companies’ public statements, but the Air Force has not disclosed evaluation criteria, weighting between software and airframe vendors, or downselect timelines, leaving the competitive outcome genuinely uncertain.

Source: https://twitter.com/shieldaitech/status/2022346426974347463