vcrouter
Start a battlePlay
Leaderboards / Market map / Battle
Market map

Defense tech: autonomous systems and battlefield software

September 13, 2026 · $0.10 for both answers and the judge

Claude Fable 5 wins

On September 13, 2026, Claude Fable 5 and Gemini 3.1 Pro each wrote a market map for the same brief, without seeing the other's answer. The judge, GPT-5.6 Luna Pro, read both twice, once in each order, and picked Claude Fable 5 both times.

Why, in the judge's words“A has sharper value-chain segmentation, stronger procurement insight, and more credible white-space calls; B relies on generic layers and overstates C2 network effects.”

The two answers

Claude Fable 5Winner
$0.079 · 27s · 1,546 tokens

Defense Tech: Autonomous Systems & Battlefield Software — Market Map

(1) Thesis

Software-defined, attritable autonomy is restructuring defense procurement away from decade-long exquisite platform programs toward venture-backed primes that sell iterating software and cheap mass — with Ukraine as the forcing function and the DoD's Replicator initiative as the institutional signal.

---

(2) Segments

A. Full-Stack "New Primes" (autonomy + hardware + software)

Companies: Anduril, Shield AI, Palantir, General Atomics (incumbent crossover), Helsing (EU) Dynamics: Winner-take-most competition to become the software-native prime; capital intensity is high but so are program-of-record payoffs (e.g., Anduril's CCA win). Palantir is software-only but functions as a prime via Maven/TITAN.

B. Attritable Drones & Loitering Munitions

Companies: AeroVironment, Skydio, Neros, Mach Industries, Firestorm Labs; less sure: Performance Drone Works Dynamics: Ukraine-driven demand for cheap, jamming-resistant, mass-producible airframes; margins compressed by commoditization, so value migrates to autonomy stacks and manufacturing throughput.

C. Maritime & Undersea Autonomy

Companies: Saronic, Saildrone, Ocius (AU); Anduril (Ghost Shark/Dive-LD, cross-segment); less sure: Blue Water Autonomy Dynamics: Fastest-growing segment on Taiwan-scenario logic (hellscape doctrine); fewer incumbents than air, so new entrants face less prime resistance. Saronic raising at multi-billion valuations to build shipyards.

D. Battlefield Software: C2, Targeting & EW

Companies: Palantir, Anduril (Lattice), Scale AI (Donovan), Second Front Systems, Epirus (counter-drone/EW hardware-software); less sure on traction: Vannevar Labs, Rebellion Defense (struggling) Dynamics: The connective-tissue layer (kill chains, sensor fusion, JADC2) — sticky, high-margin, but sales cycles gated by accreditation (ATO) and incumbent SI resistance.

E. Counter-UAS & Base Defense

Companies: Epirus, Fortem Technologies, Dedrone (acquired by Axon), Anduril (Anvil/Roadrunner), DroneShield (AU) Dynamics: Urgent demand post-Red Sea and drone incursions on US bases; fragmented across kinetic/EW/directed-energy approaches; likely consolidation as customers want layered single-vendor solutions.

---

(3) Where Value Concentrates

Segment A (Full-Stack New Primes) captures the most value. Rationale:

  • Programs of record (CCA, Replicator, IVAS follow-ons) are billion-dollar annuities; only full-stack players can bid them.
  • The autonomy software layer is the durable moat, but distribution — being the prime of record — is what monetizes it. Anduril's model (own the hardware to sell the software) has proven this.
  • Segment D has the best margins but caps out smaller; Segments B/C/E increasingly become suppliers to Segment A players, who capture integration rent.

Anduril is the presumptive category winner; Palantir wins the pure-software analog.

(4) White Space

  1. Autonomy test, evaluation & assurance infrastructure ("V&V-as-a-service"). No neutral third party exists to certify autonomous weapon behavior at scale — the DoD's bottleneck to fielding Replicator-class systems is trust, not hardware. A "UL/Underwriters Lab for lethal autonomy" with simulation-based certification would sit at a mandatory chokepoint. (Applied Intuition is adjacent from automotive but not defense-native at scale; less sure how far their defense unit has gotten.)
  2. Coalition-grade interoperability middleware. Ukraine exposed that allied systems (US, EU, indigenous Ukrainian software like Delta) can't share targeting data cleanly across classification and national boundaries. No company owns cross-national kill-chain translation — Palantir does it bespoke, but a productized, exportable-by-design NATO interop layer (built for ITAR-light release) is unclaimed and structurally hard for US primes to build.
Gemini 3.1 Pro
$0.018 · 15s · 1,510 tokens

Here is an end-to-end venture market map for Defense Tech: Autonomous Systems and Battlefield Software.

1. Investment Thesis

The defense market is being fundamentally restructured from a reliance on expensive, exquisite, hardware-centric legacy platforms to attritable, software-defined, and interoperable systems driven by commercial AI and edge-compute paradigms.

2. Market Segments & Dynamics

Segment 1: AI-Enabled Attritable Autonomy (UAS/USV/UUV)

  • Companies: Anduril, Shield AI, Skydio, Saildrone, Saronic.
  • Segment Dynamic: Hardware is increasingly viewed as a commoditized delivery mechanism; the true competitive moat is the onboard autonomy software enabling GPS-denied navigation and swarm behaviors.

Segment 2: Battlefield Operating Systems & C2 (Command & Control)

  • Companies: Palantir, Helsing, Rebellion Defense, Applied Intuition (Defense division), Primer AI.
  • Segment Dynamic: The race to build the "single pane of glass" (JADC2) that ingests disparate, multi-domain sensor feeds and uses AI to generate real-time targeting recommendations for human operators.

Segment 3: Tactical Edge Compute & Data Infrastructure

  • Companies: Shift5, Parry Labs, Edgesource, Raft, Lattice (Anduril's software, though Anduril is a prime, Lattice acts in this space).
  • Segment Dynamic: Pushing processing power directly to the sensor or vehicle to reduce latency and bandwidth reliance in highly contested, electronically jammed environments.

Segment 4: Synthetic Environments & AI Simulation

  • Companies: Improbable (Defense), Hadean, Vrgineers, Scale AI (Defense data generation), Pitch Technologies.
  • Segment Dynamic: Accelerating the deployment of autonomous systems by training models and testing wargame scenarios in high-fidelity, physics-based digital twins before real-world fielding.

3. Value Capture

Battlefield Operating Systems & C2 will capture the most value. Why: The C2 software layer acts as the "iOS of the battlefield." By integrating all sensor data and hardware into one platform, these companies benefit from extreme vendor lock-in, massive network effects, and high-margin SaaS/software economics. As they become the central nervous system of the military, they effectively commoditize the hardware and autonomous vehicles operating beneath them, forcing hardware providers to become mere "apps" on their operating system.

4. White-Space Opportunities

  • Opportunity 1: "Vanta for DoD" - Automated ATO (Authority to Operate) Infrastructure. The biggest bottleneck for defense tech is the 12-to-18-month compliance process to get software approved for classified networks. There is a massive white space for an automated compliance engine built specifically to continuously monitor and fast-track ATOs for autonomous systems and battlefield software updates.
  • Opportunity 2: Hardware-Agnostic Counter-Swarm EW (Electronic Warfare) API. While many companies build drones or counter-drone hardware, there is no dominant, open-architecture software layer dedicated solely to dynamically routing EW frequencies, spoofing tactics, and spectrum management across heterogeneous drone swarms in real-time. A purely software-defined EW orchestration layer that plugs into any existing radio or drone is currently missing.

Claude Fable 5 vs Gemini 3.1 Pro, overall

Across 30 judged battles, Claude Fable 5 won 28, Gemini 3.1 Pro won 1, and 1 were ties. On market map alone: 10 to 0.

Compare Claude Fable 5 and Gemini 3.1 Pro Judge one yourself, blind

Costs are the provider's list prices at the time of the run.