It’s Not Just Anduril and Palantir Anymore: The 2026 Defense-Tech Landscape

United States
Written By Don Bradford
It’s Not Just Anduril and Palantir Anymore: The 2026 Defense-Tech Landscape — LodiEye

It’s Not Just Anduril and Palantir Anymore: The 2026 Defense-Tech Landscape

LodiEye — June 2026

Summary

This is a follow-up to the original LodiEye analysis published February 25, 2026, Anduril and Palantir: AI-Enabled Transformation of US Defense. Between late February and early June 2026, several open questions from that piece resolved — mostly in the direction it anticipated, and faster. Anduril raised a record round and started building combat drones months ahead of schedule; Palantir posted its fastest growth since going public; the Army’s Next Generation Command and Control prototype cleared its division-scale rehearsal; and a new, much larger joint program — the Golden Dome missile shield — has pulled both companies toward the center of the most expensive weapons effort in U.S. history. Around them, defense-technology venture funding has set an all-time record, surfacing a second tier of well-capitalized startups worth tracking.

This update is organized as (1) updates to the two principal companies, (2) the status of NGC2 going into its summer test, (3) the new Golden Dome program, (4) the broader defense-tech surge and the new names within it, (5) the new hardware — drones, autonomous vessels, and weapons — these companies are now producing, and (6) a quick-reference table reconciling specific claims in the original article against current facts.

Anduril: A Record Round and Arsenal-1 Goes Live Early

The biggest change is scale. In May 2026, Anduril closed a $5 billion Series H led by returning investors Thrive Capital and Andreessen Horowitz, roughly doubling its valuation to about $61 billion — up from $30.5 billion less than a year earlier. That makes it the most valuable venture-backed defense startup in the world. The company has signaled it will spend the capital aggressively on manufacturing, research, and infrastructure, and is openly running an operating loss in 2026 (projected near $1.2 billion) as it invests ahead of revenue on large multi-year contracts.

Arsenal-1 started production ahead of schedule. The original article listed Q2 2026 as the target for first output, against the company’s own public “July 2026” guidance. In reality, Anduril began manufacturing the YFQ-44A Fury autonomous combat aircraft at the Columbus-area facility in late March 2026. The first building — roughly 775,000 square feet of production space — opened with a ~30-person “Fury Launch Team,” scaling toward about 250 employees by the end of 2026 and an eventual 4,000. At full strength (three shifts), the line is sized for roughly 150 Fury aircraft per year.

By the end of 2026, Anduril plans to stand up additional Arsenal-1 lines for the Roadrunner vertical-takeoff interceptor, the Barracuda low-cost cruise missile, and a classified platform — validating the “software-defined factory” thesis from the original piece, in which common machinery and a shared operating system let one plant switch between product types. The distributed-production picture also filled in: the Mississippi solid rocket motor factory is online and targeting 6,000 motors per year, and a roughly $1 billion campus in Long Beach and Lakewood, California is slated to open in mid-2027.

Two other developments expand Anduril’s footprint beyond the original scope. The Pentagon announced an agreement with Anduril and three other firms (CoAspire, Leidos, and Zone 5) to buy more than 10,000 low-cost cruise missiles over three years. And Anduril signed a definitive agreement to acquire ExoAnalytic Solutions, a space-sensing and missile-defense software firm — a direct move toward the Golden Dome work described below.

Palantir: Fastest Growth Since Its IPO

Palantir’s first-quarter 2026 results (reported May 4) showed the trajectory accelerating rather than leveling off. Total revenue grew about 85% year-over-year to $1.63 billion — the company’s fastest pace since its 2020 public listing — with net income roughly quadrupling to about $871 million. U.S. government revenue, the segment most relevant to the original analysis, grew 84% to $687 million, an acceleration from the prior quarter. On the strength of that quarter, Palantir raised full-year 2026 revenue guidance to roughly $7.65 billion, about a 71% annual increase.

New contract surface area. Beyond the ~$10 billion Army Enterprise Software Agreement and the expanded Maven Smart System work covered originally, Palantir added a USDA contract with a $300 million ceiling (farmland security, supply-chain resilience, fraud prevention) and deployed a “Ship OS” for the Navy that the company says compressed certain manufacturing-approval processes from about 200 hours to 15 seconds. The throughline — turning legacy government data wrangling into near-real-time decision support — is the same one the original article identified.

NGC2: From Ivy Sting to Ivy Mass, With Capstone 6 Still Ahead

The original article framed three 2026 milestones for Anduril’s NGC2 prototype: the back half of the Ivy Sting series, the Ivy Mass division event (then projected for roughly May), and Project Convergence Capstone 6 (PC-C6) in summer. The first two have now happened.

Ivy Mass (Fort Carson, early May 2026). Following the five Ivy Sting events, the 4th Infantry Division ran Ivy Mass as the culminating risk-reduction exercise — validating large-scale combat operations with NGC2 providing situational awareness, data sharing, and decision support across the formation. The Army characterized it as moving NGC2 from proof of concept to a full division equipped with commercial prototype technology in under a year. Ivy Mass was explicitly staged as preparation for PC-C6.

Project Convergence Capstone 6, at the National Training Center at Fort Irwin, remains the decisive test the original piece flagged — the point at which the 4th ID fights using NGC2 instead of legacy command-and-control systems, under heavy red-team electromagnetic and cyber pressure. That event is still ahead this summer, and the “very high risk” assessment of the integrated Anduril–Palantir battlefield communications system from the 2025 Army memo remains the open question PC-C6 is designed to answer.

Two smaller refinements since February: Anduril’s NGC2 industry team (Palantir, Microsoft, Govini, Striveworks) has added transportation-data startup Shift5 and logistics firm Rune; and the parallel Lockheed Martin track with the 25th Infantry Division has continued through its own exercise series (Lightning Surge, and a demonstration during the Balikatan 2026 exercise in the Indo-Pacific). The dual-track competition the original article described is intact.

Golden Dome: A New, Larger Joint Program

The most significant addition since February is a program the original article did not cover: Golden Dome, a space-based missile-defense shield intended to intercept ballistic, cruise, and hypersonic threats. Its budget rose by $10 billion to roughly $185 billion this spring, making it among the most expensive military technology efforts in U.S. history.

Anduril and Palantir are co-developing core software for Golden Dome, alongside Scale AI, networking firm Aalyria, and Swoop Technologies, with SpaceX contributing space-based sensing and communications and the traditional primes — Lockheed Martin, RTX, and Northrop Grumman — in lead hardware roles. In April 2026, the Space Force awarded 12 companies contracts worth up to $3.2 billion to build competing space-based interceptor prototypes, with a demonstration target of 2028; Anduril was among the awardees. The consortium reportedly aims to test integrated software this summer — the same season as PC-C6.

Golden Dome is notable for LodiEye’s readers because it crystallizes the core thesis of the original article: it explicitly pairs the new software-native entrants (Anduril, Palantir, SpaceX, Scale AI) with the legacy primes on a single, very large program, and it leans on exactly the AI command-and-control and data-fusion capabilities the earlier piece described — now applied to continental missile defense rather than the tactical edge.

The Broader Defense-Tech Surge and the New Names

The original article treated Anduril and Palantir as a near-singular ecosystem. Three months on, the more accurate picture is a rapidly widening field. Venture investment in defense, national-security, and law-enforcement startups exceeded $14.6 billion in the first months of 2026 alone — already past the prior full-year record of $9.6 billion set in 2025, and far above the roughly $1.6 billion seen in 2020. Anduril’s $5 billion round is the single largest contributor, but a second tier of well-funded companies has emerged.

U.S. Defense-Tech Venture Funding by Year ($ billions)

Source: Crunchbase data as reported June 2026. 2026 figure is year-to-date; funding held relatively stable across 2022–2024 before the sharp 2025–2026 increase.

The trend is not just larger totals but a transition from fundraising to fielding. The following companies have moved from prototypes toward signed production programs, and are worth tracking alongside the two principals.

Company Focus Recent Funding / Status (2025–2026)
Shield AI Autonomous flight software (Hivemind) for GPS-denied environments; tested on F-16s and Collaborative Combat Aircraft $2 billion Series G (March 2026) at a ~$12.7 billion valuation; acquiring simulation maker Aechelon
Saronic Autonomous surface vessels (naval drones); building an autonomous shipyard $1.75 billion Series D (March 2026) at a ~$9.25 billion valuation
Castelion Affordable, mass-produced hypersonic missiles $350 million Series B (December 2025); new New Mexico production facility
CHAOS Industries Counter-drone radar and sensing Valued around $4.5 billion; Series D in late 2025
Epirus Directed-energy (high-power microwave) counter-drone systems Series D in the $250 million range
Mach Industries Autonomous drone and strike systems $300 million Series C
Scale AI / Aalyria / Swoop AI data integration; laser/optical networking; defense software — all on the Golden Dome software team Active 2026 program participants
Helsing (Europe) AI software for defense; Europe’s leading defense-AI firm and a non-U.S. analogue Valued around €12 billion
Space-adjacent (True Anomaly, Vast, Sierra Space, Apex, K2 Space) Orbital infrastructure with national-security applications — a fast-growing sub-vertical feeding programs like Golden Dome Among the largest 2026 defense-related recipients

The shift underneath the numbers: 2025 was largely about invention and fundraising; 2026 is about execution — converting facilities into repeatable production output. Manufacturing scale, not just software, has become the competitive battleground, which is precisely the bet Arsenal-1 represents. There are even early signs of exits: AI drone company Swarmer went public this year, with shares jumping more than 500% on their first trading day.

New Hardware: Drones, Autonomous Vessels, and Weapons

The most concrete changes since February are in the hardware itself. Anduril has moved several systems from announcement to signed production agreements, and the newer entrants have fielded or detailed platforms across the air, sea, and counter-drone domains. The systems below are those that are new, newly contracted, or newly detailed since the original article. (Note: the Department of Defense now also operates under the designation “Department of War”; both names appear in current contracting announcements, and “the Pentagon” is used here as a neutral shorthand.)

Anduril: Modular “Families” Across Air, Sea, and Strike

Anduril’s portfolio has consolidated into families of modular systems, in which a common airframe or hull spawns both sensing and strike variants — the latter typically carrying an “M” (munition) designation. The unifying layer remains Lattice OS, which connects every platform into a single operational picture.

System Domain / Type What’s New or Notable
Fury (YFQ-44A) Air — collaborative combat aircraft Group 5 autonomous “loyal wingman”; in production at Arsenal-1 since late March 2026; an Air Force Collaborative Combat Aircraft follow-on production decision is expected later in 2026
Barracuda-100 / -250 / -500 Air — low-cost autonomous air vehicles / cruise missiles Mass-producible family (about 70% commercial components); “M” variants deliver kinetic effects; the Barracuda-250 is sized to fit inside fifth-generation fighter internal weapons bays
Surface-Launched Barracuda-500M Weapon — long-range precision strike May 13, 2026 framework agreement with the Pentagon for at least 3,000 missiles and 60-plus containerized launchers over three years, deliveries from 2027; range exceeding 500 nautical miles; part of the Low-Cost Containerized Missiles program (10,000-plus missiles total across Anduril, CoAspire, Leidos, and Zone 5)
Roadrunner / Roadrunner-M Air — reusable interceptor Vertical-takeoff air-defense interceptor for counter-drone and counter-cruise-missile missions; recoverable and re-flyable if it does not engage
Copperhead-100 / -500 (and -500M) Maritime — autonomous underwater vehicles / munitions New undersea family; non-“M” versions are sensing/ISR AUVs, while the “M” versions (e.g., Copperhead-500M, first public footage April 2026) are torpedo-like underwater munitions; Anduril was selected by the Defense Innovation Unit for the Combat Autonomous Marine Platform (CAMP) project in March 2026
Ghost Shark Maritime — extra-large UUV (XLUUV) Flagship AUKUS program; Australia’s roughly $1.12 billion fleet deal is being built in-country, a model for distributed allied production
Anvil / Anvil-M Counter-drone — kinetic interceptor Drone-ramming interceptor; the Anvil-M variant adds a high-explosive warhead for harder targets
Pulsar Electronic warfare Family of software-defined EW systems that detect and jam drone command-and-control links across the counter-UAS fight
EagleEye Soldier systems — mixed reality Augmented-reality headset on the path to succeed the IVAS program Anduril absorbed from Microsoft; roughly 100 units targeted for Army delivery in 2026

European production path: Anduril and Germany’s Rheinmetall announced a strategic partnership to co-develop European variants of the Barracuda and Fury, integrating them into Rheinmetall’s “Battlesuite” digital framework, alongside exploration of European solid rocket motor production — aimed at the roughly €800 billion in projected European defense spending through 2027.

Saronic: A Graduated Family of Autonomous Boats

Saronic now fields a full line of autonomous surface vessels (ASVs), from a six-foot scout to a 150-foot-plus medium unmanned vessel, all running its Echelon software — which lets a single sailor supervise one boat or a swarm. The company is explicitly pitching American shipbuilding capacity as a strategic asset against China.

Vessel Size Role / Status
Spyglass ~6 ft Small ISR / scout ASV
Cutlass ~14 ft Short-range maritime security and ISR
Corsair 24 ft Flagship multi-mission ASV: ~1,000 lb payload, 1,000-plus nautical-mile range, under $1 million per unit; $392 million Navy production contract (December 2025); in June 2026 a Navy-operated Corsair carried out the first publicly reported personnel rescue by an unmanned surface vessel, recovering two Army aviators off Oman
Mirage ~52 ft Force-multiplier ASV: ~3,500 lb payload, ~2,500 nautical-mile range
Cipher ~60 ft Larger multi-mission autonomous vessel
Marauder ~150–180 ft (MUSV) Medium Unmanned Surface Vessel with ~150-metric-ton payload (ISO-container modular); first hull launched May 31, 2026 at the Franklin, Louisiana shipyard, which is targeting up to 20 per year by end of 2026; selected for the Navy’s MUSV Marketplace, with at-sea demonstrations June–October 2026

The strategic backdrop: the Navy has signaled intent to make roughly half its surface fleet unmanned, and Saronic says it built the first Marauder hull in under a year — a pace it characterizes as the fastest U.S. shipbuilding since World War II.

Shield AI: From ISR Drone to Autonomous Fighter

Shield AI’s hardware is anchored by the V-BAT (designated MQ-35), a single-engine ducted-fan drone that takes off and lands vertically from ships or confined ground sites and runs the company’s Hivemind autonomy — which uses visual navigation to operate in GPS- and communications-denied conditions and to fly in coordinated swarms. The V-BAT offers Group 4/5 capability in a Group 3 package, and recent customers include the Indian Army (a roughly $35 million emergency procurement of V-BAT plus a Hivemind license, January 2026), Japan’s Maritime Self-Defense Force, and a U.S. Navy contractor-owned ISR program.

X-BAT — the higher-end bet: Shield AI is developing the X-BAT, an AI-piloted, runway-independent vertical-takeoff combat jet built around Hivemind and a human-on-the-loop control model. In about 18 months the program has moved through wind-tunnel, pole, and engine testing, with a structural airframe in fabrication; first vertical-takeoff flights are scheduled for 2026 and mission-capable status is projected for 2028. Shield AI was also selected to integrate Hivemind onto the Low-Cost Uncrewed Combat Attack System (LUCAS).

Counter-Drone Directed Energy and Low-Cost Hypersonics

Two newer companies fill capability gaps the original article only gestured at — the cheap-drone-swarm problem and affordable long-range strike.

Epirus — Leonidas (high-power microwave). Leonidas is a solid-state, software-defined directed-energy system built on gallium-nitride semiconductors and a phased-array architecture; rather than firing an interceptor, it emits a beam of high-power microwave energy that overloads the electronics of drones across a wide arc. In an August 2025 live-fire demonstration it downed 61 of 61 drones, including a 49-drone swarm in a single pulse, and in December 2025 it became the first electronic-warfare system to defeat a fiber-optic-guided FPV drone — a threat immune to conventional radio jamming. The company markets it as “pennies per kill” with an effectively unlimited magazine. Leonidas now comes in multiple form factors: a vehicle kit, a sub-50-pound backpack-sized pod (mountable on drones, vehicles, or helicopters — including Anduril’s Roadrunner), and autonomous ground-vehicle versions developed with General Dynamics Land Systems and Kodiak AI.

Castelion — Blackbeard (low-cost hypersonic). Castelion signed a production framework agreement with the Pentagon for its Blackbeard hypersonic strike missile, establishing a guaranteed minimum of 500 missiles per year once testing is validated, with a pathway to thousands more under the “Arsenal of Freedom” initiative. The company describes it as the first production arrangement of its kind for a low-cost hypersonic weapon. Mach Industries, similarly, is developing low-cost autonomous strike drones and loitering munitions aimed at the same mass-affordable-firepower thesis.

The common thread: nearly every new system above is designed around cost and volume rather than exquisite per-unit capability — cheap interceptors, container-launched missiles, swarming boats, and energy weapons that cost little per shot. It is the same “mass over exquisite” logic the original article attributed to Arsenal-1, now visible across the whole field and increasingly backed by signed production contracts rather than prototypes.

Reconciling the Original Article With Current Facts

For readers returning from the February piece, the table below maps specific forward-looking claims to their current status.

Original Claim (Feb 25, 2026) Status as of June 2026
Arsenal-1 to begin YFQ-44A production “around Q2 2026” Confirmed — and earlier: Fury production began in late March 2026, ahead of the company’s own July target
Ivy Mass full-division event “approx. May 2026” Confirmed — ran at Fort Carson in early May 2026 as the culminating risk-reduction exercise
Project Convergence Capstone 6 in summer 2026 as the key inflection point Still ahead — scheduled this summer at Fort Irwin; remains the decisive operational test
Anduril valued in the low tens of billions (prior $30.5 billion round referenced) Updated — ~$61 billion after a $5 billion Series H in May 2026
Anduril–Palantir consortium focused on tactical C2 and battlefield comms Expanded — now also co-developing core software for the ~$185 billion Golden Dome missile shield
“Very high risk” Army memo on integrated battlefield communications Unresolved — PC-C6 this summer is the intended stress test
Ecosystem framed primarily around two companies Broadened — record $14.6 billion in 2026 defense-tech venture funding and a clear second tier (Shield AI, Saronic, Castelion, and others)

What to Watch Next

  • Project Convergence Capstone 6 (summer 2026): the first division-scale, contested-environment test of NGC2 — the clearest read yet on whether the architecture holds under jamming and cyber pressure.
  • Arsenal-1 production ramp: whether the Roadrunner, Barracuda, and classified lines stand up by year-end as promised, and whether the “software-defined factory” delivers the switching speed it claims.
  • Air Force CCA decision: a follow-on production contract for the YFQ-44 Fury, expected later this year, would sharply change Arsenal-1’s output.
  • Golden Dome software test (this summer) and the 2028 interceptor demonstration: early indicators of whether the new entrants can deliver on a flagship, multi-prime program.
  • Funding-to-execution transition: with capital records broken, the 2026–2027 question is which firms convert money into fielded, contracted production — and which consolidate or exit.

LodiEye is the investigative research arm of Lodi411.com, a citizen-run civic data and transparency platform serving Lodi, California and San Joaquin County. LodiEye is not a traditional news outlet. It does not employ professional journalists or reporters, and the people behind it do not hold journalism degrees or have professional newsroom experience. LodiEye is best understood as civic research and analysis — not peer journalism — and is not a substitute for the local and regional news organizations that do this work professionally. For traditional reporting on Lodi, San Joaquin County, and the broader region, readers are encouraged to consult the Lodi News-Sentinel, Stocktonia, The Sacramento Bee, CalMatters, and other established news outlets staffed by credentialed journalists.

This LodiEye update was produced using artificial intelligence tools under the direction and review of the founder. Lodi411 uses multiple AI platforms in its research and publication workflow, including Anthropic’s Claude (primarily Opus and Sonnet models) and Perplexity AI across a variety of large language models offered by each. For this update, Anthropic’s Claude (Opus) performed the live web research and drafting; the tools were used in the following capacities:

Source Discovery: AI-assisted web search and retrieval identified roughly three dozen sources published between February and June 2026, spanning defense trade press (Breaking Defense, Defense One, Reuters, CNBC, Naval News), primary company announcements (Anduril, Saronic, Shield AI, Epirus, Castelion), U.S. Army releases (army.mil), and corporate financial filings (Palantir SEC Form 8-K and 10-Q).

Credibility Validation: Key figures — funding amounts, contract values, production dates, and platform specifications — were cross-referenced across multiple independent sources, prioritizing primary company and government releases and SEC filings, then established defense trade press, then aggregators. Figures that varied by source or definition (for example, the Marauder vessel’s length and the defense-tech funding totals) were flagged in the text rather than presented as settled.

Analysis and Synthesis: Claude organized three-plus months of developments against the original February article, building the reconciliation table that maps each forward-looking claim to its current status and grouping the new hardware by domain (air, maritime, counter-drone and directed energy, and strike).

Presentation: Claude assisted in drafting and structuring the update in Lodi411’s HTML format, including the defense-tech funding chart and the comparison tables.

Final Review: The completed draft was reviewed for factual consistency, source attribution, and logical coherence; cross-checking each key claim against multiple independent sources is the primary error-reduction mechanism. All editorial judgments, analytical conclusions, and the decision to publish were made by the founder.

Lodi411/LodiEye believes transparency about AI use serves both readers and the broader information ecosystem. Readers who spot errors are encouraged to write editor@lodi411.com so corrections can be made.

References

For questions or follow-up, contact the Lodi411 editorial team at editor@lodi411.com.

Don Bradford
Previous

Silencing the Horns: What a Downtown Lodi Quiet Zone Would Take, and What It Could Mean
Next

Lodi City Council Agenda - June 17, 2026