Defense AI and the $10B Neo-Prime Consortium: How Silicon Valley Is Disrupting the Pentagon (feat. Palantir, Anduril, Golden Dome)

In December 2024, six Silicon Valley companies convened in secret.

Palantir, Anduril, SpaceX, OpenAI, Scale AI, and Saronic. The alliance they formed: the Neo-Prime Consortium.

Their target market: the $850B U.S. defense budget. In plain terms, this was Silicon Valley’s declaration of war on Lockheed Martin and Boeing’s decades-long monopoly over American defense.

The first article in this series dissected Palantir’s financials. The second deconstructed its Ontology technology. This installment tracks how Palantir makes money on the battlefield, and why a new force called “Neo-Prime” threatens the legacy defense giants.

The core question: “Can software beat a tank?”

The Neo-Prime Consortium: Who Joined, and Why

The 6 Members and Their Roles

The name itself is significant. In the U.S. defense industry, “Prime” refers to the Big 5 — Lockheed Martin, Boeing, Raytheon, Northrop Grumman, and General Dynamics. Neo-Prime is a declaration: “We are the new Primes.”

The member roles reveal the strategy. Palantir handles data analytics and battlefield command platforms. Anduril covers autonomous weapons and drones. SpaceX provides launch vehicles and satellite networks. OpenAI supplies AI models. Scale AI manages AI training data. Saronic builds unmanned surface vessels.

Six companies lock together like puzzle pieces completing a single weapons system.

Where traditional defense consolidated everything within a single massive corporation, Neo-Prime assembles the best in each domain into a “coalition force.”

The Decisive Difference from Legacy Defense

The fundamental gap:

DimensionLegacy Primes (Lockheed Martin et al.)Neo-Prime (Palantir, Anduril et al.)
Core CompetencyHardware (fighters, missiles, ships)Software + AI + autonomous systems
Development CycleYears to decades (F-35: 23 years)Weeks to months (Maven: 30-day NATO deployment)
Revenue ModelCost-plus (cost + guaranteed margin)Fixed-price or performance-based contracts
Revenue Growth3-5% (Lockheed 5%, Boeing defense in the red)50-100%+ (Anduril 2x+, Palantir defense 66%)
Procurement ModelSingle large contractsModular, multi-vendor integration
Signature ContractF-35: $1.7T cumulativeArmy $10B, Maven $1.3B

The numbers are stark. Lockheed Martin’s 2025 revenue was $75B, #1 in the industry — but growing at just 5%.

Anduril’s defense revenue hit approximately $950M, more than doubling year-over-year. It debuted at #93 on the Defense News Top 100. Palantir ranked #70 with approximately $1.6B in defense revenue.

Boeing’s defense division is in worse shape — $4.9B in cumulative losses for 2024, a historic low driven by chronic cost overruns in fighter jet and satellite programs. Boeing is proving the limits of hardware-centric defense in real-time.

The structural problem: cost-plus contracting. Legacy defense charges the government cost plus a guaranteed margin. When costs balloon, revenue actually increases — a perverse incentive. The F-35 fighter’s 23-year development timeline and $1.7T cumulative cost are direct products of this structure.

Neo-Prime rejects this model. Software has near-zero marginal replication cost, enabling fixed-price or performance-based contracts. This is more efficient from both taxpayer and military perspectives.

U.S. Army CIO Leo Garciga, upon signing the $10B contract with Palantir in July 2025, called it “a harbinger of software procurement.” The deal consolidated 75 separate contracts into a single framework for up to 10 years.

DEFENSE AI MARKET OVERVIEW

$10B

Army Enterprise Contract

$175B

Golden Dome Budget Estimate

$1.3B

Maven Smart System

30 sec

Ukraine Targeting Time

Golden Dome: The $175B Missile Defense Ambition

From Space to Ground

Neo-Prime’s first major proving ground is Golden Dome — a missile defense system that extends Israel’s Iron Dome concept to the entire U.S. mainland.

On January 27, 2025, President Trump signed an executive order committing $175B over three years to build a “next-generation missile defense shield.”

The strategic imperative: China and Russia have fielded hypersonic missiles. Existing missile defense systems are effectively unable to intercept them.

Speeds of Mach 5-10 (6,000-12,000 km/h) leave zero time for human decision-making. AI must automate the entire chain from detection to intercept.

Congress approved $24.4B — $18.8B for next-gen missile defense and $5.9B for homeland defense. But the gap to the $175B target remains substantial.

Independent cost estimates are more sobering. The CBO projects $500B. The American Enterprise Institute (AEI) estimates up to $3.6T through 2045.

SpaceX plays a pivotal role in this architecture — deploying 400-1,000 sensor satellites and approximately 200 interceptor satellites in orbit.

The paradigm shift: if traditional missile defense was “catching a ball with a baseball glove,” Golden Dome proposes “stopping the pitcher on the mound before the throw.”

The obstacle is funding. As of February 2026, the OMB (Office of Management and Budget) has blocked Golden Dome budget execution. The executive order is signed, but actual funding remains frozen. Reaching $175B requires simultaneous Congressional and OMB approval — a process that could take years.

TheByteDive
The Neo-Prime Consortium is redefining defense through software from satellites to ground systems | Photo: Unsplash

Battle-Proven AI

Ukraine: From Days to 30 Seconds

The evidence that Neo-Prime technology is not theoretical lies on the Ukrainian battlefield.

Palantir’s most impressive achievement: compressing the kill chain — the time from target detection to strike — from “days” to “approximately 30 seconds.” In warfare, the difference between 30 seconds and days is the difference between life and death.

The enabling technology: MetaConstellation. A system that dynamically combines commercial satellite constellations for real-time battlefield surveillance without dedicated military reconnaissance satellites.

In January 2026, the Ukrainian government launched Brave1 Dataroom — a program opening real combat data to AI companies for model training.

Why this matters: AI feeds on data. “Real combat data” is the rarest and most expensive data in the world.

Simulation and live combat are fundamentally different. The real-world combat data Ukraine has accumulated during the war is an asset no defense company can purchase at any price.

Ukraine and the UK went further, signing a weapons data exchange agreement. This is the new form of military alliance in the AI era. Previously, nations shared weapons. Now they share data.

Army $10B + Maven $1.3B

Technology validated in Ukraine is being re-imported to the U.S. military.

Maven Smart System is the flagship example. Originally a $480M contract, it expanded with an additional $795M to total $1.3B. Currently deployed across 10 Combatant Commands, with user count doubling since January 2025 to exceed 20,000.

Maven’s target: achieve “100% machine-generated” intelligence dissemination by June 2026.

Not human analysts processing intelligence manually — AI automatically analyzing satellite, drone, and communications data and delivering it to commanders.

The NATO expansion is particularly noteworthy. From contract signing on March 25, 2025, the system was operational in just 30 days. The shortest procurement-to-deployment timeline in NATO history.

Traditional defense procurement averages 5-7 years. The speed differential of software-defined defense is stark.

The Army Enterprise $10B contract (signed July 31, 2025) consolidated 75 individual software contracts into a single framework, up to 10 years — restructuring software procurement across the entire U.S. Army.

TITAN (Tactical Intelligence Targeting Access Node) adds another datapoint. A $178M contract for 10 ground intelligence analysis systems. The first 2 units were delivered in March 2025, with full delivery targeted for 2026.

The common thread: every one of these contracts replaces projects that legacy defense contractors attempted for years and either failed or delayed.

Palantir’s defense revenue growing 66% in 2025 to $1.6B traces directly to this pattern. Maven, TITAN, Army $10B — each one carving into the legacy Primes’ market share.

Legacy Defense vs. Software-Defined Defense

Legacy Defense (Prime)

  • Hardware-centric (tanks, fighters, ships)
  • 10-20 year development cycles
  • Routine cost overruns
  • Lockheed Martin, Boeing, Raytheon

Software-Defined Defense (Neo Prime)

Traditional Defense (Prime)

Software Defense (Neo Prime)

  • AI & Software-Centric Platform
  • Deployment in weeks to months
  • Low fixed costs, rapid scaling
  • Palantir · Anduril · Shield AI
  • Hardware-centric (tanks, fighters, ships)
  • Development cycle 10~20 years
  • Cost overruns routine
  • Lockheed Martin · Boeing · Raytheon

개발 주기 10~20년

  • 비용 초과 일상적
  • 록히드마틴·보잉·레이시온
    • AI and software-centric platforms
    • Weeks-to-months deployment
    • Low fixed costs, rapid scaling
    • Palantir, Anduril, Shield AI

    Anduril: Arsenal-1 and the Age of Autonomous Weapons

    Within Neo-Prime, if Palantir is the brain, Anduril is the muscle. Founded by Palmer Luckey — who sold the Oculus VR headset to Facebook before pivoting to defense.

    In 2025, Anduril raised $2.5B in a Series G round at a $30.5B valuation. Founders Fund led the round, which was 8-10x oversubscribed. No defense startup has ever commanded this kind of valuation.

    Anduril’s Arsenal-1 is a $1B autonomous weapons factory under construction in Ohio. At 5 million square feet (approximately 465,000 square meters), production is scheduled to begin in summer 2026.

    The first mass-production item: the YFQ-44A Fury autonomous combat aircraft. A pilotless, AI-controlled fighter. Anduril claims it can be mass-produced at 1/10 the cost of manned fighters.

    Losing an aircraft means zero pilot casualties. The cost exposure drops dramatically.

    Anduril also acquired the U.S. Army’s troubled IVAS AR headset contract ($22B) from Microsoft, which had repeatedly failed field testing. Anduril plans to redesign the system from scratch.

    The Question for Global Defense Industries

    Global defense industries face the same inflection point. The world’s defense exporters — many of whom rely heavily on hardware sales — must reckon with Neo-Prime’s challenge: “How long can hardware without software remain competitive?”

    The Ukraine battlefield proved that advanced tanks can be rendered helpless by drone swarms. $5M Russian tanks are destroyed daily by $500 drones.

    The value proposition of weapons is migrating from “unit cost” to “network-connected system intelligence.”

    Software and AI investment is urgent for every defense manufacturer still riding a hardware export boom.

    Meanwhile, Anduril is building a $1B factory, and Palantir onboarded all of NATO onto its platform in 30 days.

    At the individual career level, defense is no longer a conservative government domain. Silicon Valley’s top talent is gravitating toward defense.

    Palmer Luckey has repeatedly stated: “The best technology problems of the 21st century are in defense.” For software engineers and AI researchers, defense represents a new career frontier.

    The strategic question: will defense industries rest on their current hardware export success, or will they pivot to Software-Defined Defense? The answer to this question will determine global defense positioning for the next decade.

    One-Line Take. The battle for the $850B U.S. defense budget is no longer Lockheed Martin vs. Boeing. Palantir and Anduril’s Neo-Prime consortium is proving on the battlefield that “software beats tanks.” The clock is running out on legacy defense.

    Actionable Insight. The Software-Defined transition is accelerating not only in defense but across every legacy industry. The question worth asking: “Does our company’s core value reside in hardware or software?”

    Sources

    1. Palantir Deep Dive: 70% Revenue Growth and the Reality Behind $313B Valuation (TheByteDive, 2026-02-19)
    1. Palantir Ontology Deep Dive: Why Palantir Spent 20 Years Building the Ontology (TheByteDive, 2026-02-25)
    1. Palantir Q4 2025 Earnings — Army $10B, Maven $1.3B (Palantir IR)
    1. Anduril Series G $30.5B Valuation (TechCrunch, 2025)
    1. Golden Dome Executive Order (White House, 2025-01-27)
    1. Ukraine Brave1 Dataroom (Ukraine MoD, 2026-01)
    1. Defense News Top 100 Rankings (Defense News, 2025)
    1. CBO/AEI Golden Dome Cost Estimates (2025)

    Palantir Deep Dive Series (6 Parts)

    FAQ

    What is the Neo-Prime Consortium?

    A defense AI consortium formed by Palantir, Anduril, SpaceX, OpenAI, Scale AI, and Saronic. Its goal is to redesign U.S. military systems with software-centric architecture, replacing legacy hardware-focused defense contractors.

    What is the Golden Dome project?

    A missile defense system proposed by the Trump administration with an estimated budget exceeding $175B. It extends Israel’s Iron Dome concept to the U.S. mainland, with Palantir and Anduril as core technology providers.

    How large is Palantir’s defense revenue?

    Palantir’s government revenue reached $1.58B in 2025 (+45% YoY). The Maven Smart System, its largest U.S. military AI contract, stands at $1.3B.

    What is Anduril’s Arsenal-1 factory?

    An autonomous weapons manufacturing facility under construction in Ohio. It will mass-produce autonomous weapons systems, converting the traditional low-volume, high-cost defense model to high-volume, software-centric production.

    How does this affect global defense industries?

    Defense exporters worldwide remain hardware-centric. Without transitioning to software-defined defense, hardware-only growth will face structural limits as battlefield value migrates from unit cost to network intelligence.

    Related Articles

    This article is for informational purposes only and does not constitute a recommendation to buy or sell any specific security. Investment decisions should be made at your own discretion and risk.

    Found this helpful?

    ☕ Buy me a coffee

    Leave a Comment