Air Force Working with Collins, Shield AI to Build Autonomy Software for CCAs

The Air Force confirmed Collins Aerospace and Shield AI as the two mission-autonomy software providers for its Collaborative Combat Aircraft (CCA) program and integrated both into the Autonomy Government Reference Architecture (A-GRA), enabling vendor-agnostic deployment of algorithms. Collins’ Sidekick has already been flight-tested on General Atomics’ YFQ-42A—controlling the aircraft for more than four hours and executing human-provided commands with high accuracy—while Shield AI expects near-term flight tests of its Hivemind software on Anduril’s YFQ-44A. The moves formalize a two-track vendor pairing (Collins/General Atomics vs. Shield/Anduril), accelerate commercialization of mission autonomy in combat aviation, and could influence procurement dynamics and equity prospects across prime defense contractors and Silicon Valley defense startups.

Analysis

Market structure: The confirmed pairing (Collins/RTX with General Atomics; Shield AI with Anduril) crystallizes winners — RTX (Collins) and AI/autonomy software specialists — and keeps General Atomics/Anduril as prime platform beneficiaries. Expect RTX to capture incremental revenue from integration and sustainment; estimate a realistic 1–3% revenue lift for RTX over 12–24 months if CCA flight-test cadence continues. Open A‑GRA standard lowers vendor lock‑in, pressuring proprietary avionics margins and accelerating competitive bidding among software suppliers.

Risk assessment: Tail risks include an autonomy-related incident or Congressional/regulatory push (export controls, certification delays) that could pause flight testing for 3–18 months and slash near‑term program revenues by >30% for smaller integrators. In the immediate term (days–weeks) market reaction will be muted; the next 3–12 months of test data and FY26 DoD funding cycles are the critical windows. Hidden dependency: sustained demand hinges on budget allocations and allied exportability; a drop >10% in DoD CCA budget guidance would materially change economics.

Trade implications: Tactical trades favor public defense primes with software/integration exposure and semiconductor/AI compute providers. Consider a 2–3% long in RTX (RTX) and 1–2% long in NVDA (NVDA) or AMD (AMD) to capture compute demand for autonomy. Use 6–12 month call spreads on RTX to cap cost; hedges: buy 3–6 month puts on a smaller systems supplier (e.g., HON) if you hold it, or buy a $100k notional 12‑month tail put on the defense basket to guard against regulatory shocks.

Contrarian angles: The market underestimates commoditization risk — A‑GRA could turn mission autonomy into a repeatable, low‑margin commodity within 3–5 years, benefiting platform-agnostic integrators but squeezing pure‑play autonomy valuations. Historical parallel: early UAV hype produced long delays between demonstrations and durable revenue (multi-year lags). Unintended consequences include increased cyber-risk premiums and export constraints that could bifurcate US vs allied supply chains, advantaging firms with strong DoD relationships and physical manufacture onshore.

Key Decisions for Investors

• Establish a 2–3% long position in Raytheon Technologies (RTX) over the next 2–8 weeks, targeting +15–25% upside in 6–18 months contingent on continued flight-test milestones; use a 9–12 month call spread (buy 12‑month ATM calls, sell ~20% OTM calls) to reduce premium.
• Allocate 1–2% to semiconductor/AI names (NVDA or AMD) to capture edge/ground-station compute demand; prefer 6–12 month call options (buy OTM 3–6 month calls rolling to LEAPS on strong earnings) and take profits on +25% move.
• Initiate a pair trade: long RTX (1–2%) / short Boeing (BA) (1%) for 3–12 months — rationale: RTX gains direct integration revenue while BA carries greater program execution risk; tighten stops at 10% adverse move.
• Buy 3–12 month protective puts (or a small put‑spread) on a concentrated defense-software/sensor small‑cap exposure equal to 0.5–1% portfolio notional to hedge a regulatory or incident-driven program pause (event risk horizon 3–18 months).
• Reduce cyclicals/exposure to non‑defense avionics suppliers by 1–3% and reallocate into defense primes (RTX, NOC, LMT) over the next quarter; re-evaluate if DoD FY26 guidance cuts CCA funding by >10% or if flight tests are grounded for >3 months.