Anduril and PGZ to Build 5,000 Cruise Missiles Annually in Poland

The Deal

Poland will build several thousand Barracuda-500M cruise missiles a year on its own soil. The agreement signed July 6 at Wojskowe Zakłady Lotnicze No. 2 in Bydgoszcz makes PGZ the exclusive European licensee for Anduril's surface-launched variant, marking the first concrete production line for the missile outside the United States.

Prime Minister Donald Tusk, Defence Minister Władysław Kosiniak-Kamysz, and State Assets Minister Wojciech Balczun attended the ceremony. Their presence signaled the strategic weight Warsaw assigns to a program that moves the country from buyer to producer of long-range precision fires.

The deal expands a memorandum of understanding the two sides struck in October 2025. That MoU framed the partnership as technology transfer and joint development. The July agreement converts it into a phased industrial plan: assembly capability first, then full-rate production, then a supply chain that progressively shifts to Polish and European subcontractors. The end goal is a missile that qualifies as majority European-made under the EU's SAFE programme.

"We will be able, within a short period of time, to produce and deliver several thousand low-cost yet technologically advanced autonomous long-range cruise missiles," said Adam Leszkiewicz, PGZ president. He called the contract a breakthrough for national defence and for the Polish defence industry's competencies.

Brian Moran, Anduril's vice president for Europe, framed it in deterrence terms: "The future of deterrence belongs to nations that can produce advanced capabilities quickly, affordably, and in the quantities required to sustain and win a fight." He described the Bydgoszcz line as a step toward a European industrial base that can replenish precision fires "at the speed of relevance."

The missile carries a 100-pound (45 kg) warhead beyond 500 nautical miles (925+ km). Anduril designed it for mass production from the start, featuring simplified architecture, common tooling, and a resilient supply chain. The company claims 50 percent less build time, 95 percent fewer tools, and 50 percent fewer parts than legacy cruise missiles. Unit cost at rate: under $150,000.

Anduril signed a broader European framework with Rheinmetall in June 2025 covering Barracuda and Fury co-development integrated into the German firm's Battlesuite ecosystem. That agreement set no production location. The PGZ deal is the first to pour concrete.

Poland is not the first licensee, as Taiwan preceded it. But the Bydgoszcz line marks the first time a NATO European ally will manufacture the missile domestically, with exclusive rights to adapt the design and software for Polish and wider European requirements. The U.S. Department of Defense's own framework agreement with Anduril in May 2026 to scale SLB-500M production underscores the missile's growing role in allied stockpiles.

Poland Becomes Producer, Not Just Buyer

Poland has become the alliance's spending outlier. In 2024 it devoted 4.07 percent of GDP to defense, the highest share in NATO, and directed nearly half that budget to major equipment, including R&D. Only Estonia and Latvia approach its GDP ratio; no other ally crosses the 3 percent threshold. The NATO Secretary General's 2024 report shows European Allies and Canada collectively spent $486 billion, a 19.4 percent real-terms jump from 2023, yet the same document warns that adversaries have been gearing up for competition, coercion and potentially conflict, and that the transatlantic industrial base must "exponentially accelerate" to keep pace.

The spending surge masks a hollowed-out magazine. CSIS analysis of the 2026 Iran campaign found U.S. inventories of Tomahawk, THAAD, and Patriot interceptors will take three or more years to return to prewar levels at current delivery rates. Standard Missiles need roughly two years. The U.S. FY 2027 budget requests $331 billion for munitions, a fourfold increase over recent annual procurement, but Defense Secretary Hegseth acknowledged to Senator Kelly that replenishment timelines run "months and years" depending on the system. Every allied order now competes with that U.S. replenishment queue.

According to Senator Mark Kelly's press release, the U.S. FY 2027 budget puts $120 billion into the defense industrial base and $48.8 billion into critical minerals and domestic supply chains, but also frames allied production as a way to "meet our own requirements while encouraging them to increase their defense investment and build up their own forces."

NATO's structural response is the Industrial Capacity Expansion Pledge and an updated Defence Production Action Plan. The Alliance identified EUR 34 billion in uncontracted land and air defense munitions requirements; Allies agreed to buy nearly half as a consortium rather than individually, sending industry a sustainable demand signal. Six new High Visibility Projects launched in 2024 include multinational procurement of indirectly fired munitions and a test-and-certification framework for interchangeability, the first time NATO has built a repeatable, voluntary cycle for aggregating demand across borders.

The political driver is burden sharing codified. The 2026 U.S. National Defense Strategy sets a new global standard of 5 percent of GDP for allies, explicitly stating that while the U.S. prioritizes homeland defense and deterring China, allies must take the lead against threats in their own regions with more limited U.S. support. The FY 2027 budget puts $120 billion into the defense industrial base and $48.8 billion into critical minerals and domestic supply chains, but also frames allied production as a way to "meet our own requirements while encouraging them to increase their defense investment and build up their own forces."

Poland sits at the geographic center of that logic. It hosts a Forward Land Forces battlegroup, the newest U.S. Aegis Ashore ballistic missile defense site (declared operational at Redzikowo in July 2024), and the NATO-Ukraine Joint Analysis, Training and Education Centre in Bydgoszcz. Its navy is building three Miecznik-class frigates and six Kormoran II mine countermeasure vessels; it has begun analyzing foreign submarine designs for a potential four-boat program. The 2024 Dragon exercise brought 20,000 troops from nine allies to test responses to threats against Polish territory, including complex river crossings and high-intensity live fire.

Anduril's licensed Barracuda-500M line with PGZ fits the gap the consortium model cannot yet fill: a cruise missile designed for software-defined manufacturing, produced on NATO soil by a non-traditional prime that moves from digital twin to metal in months rather than years. The Barracuda program was built to sidestep the tooling and workforce bottlenecks that keep Tomahawk and JASSM lines at a fraction of surge capacity. If the Polish line proves the model (digital thread, automated test, Lattice integration), it becomes a template for other eastern-flank states needing sovereign deep-strike magazines without waiting on U.S. allocation queues.

Inside the Factory

Anduril's Barracuda line was designed from the start for hyper-scale production, with 90 percent parts commonality across variants and a single airframe configurable for air, ground, or surface launch. The Polish factory takes that design philosophy and applies it to a sovereign European production line built with PGZ, the state-owned Polish Armaments Group that operates more than 50 subsidiaries from its Radom headquarters.

The manufacturing workflow mirrors Anduril's U.S. approach: model-based systems engineering defines every interface, common hardware modules (avionics, actuators, power systems, propulsion) move through standardized test cells, and a digital twin tracks each unit from raw material to acceptance. Where traditional primes ship finished missiles, Anduril ships a production system: tooling packages, test software, work instructions, and the Lattice autonomy layer that runs automated end-of-line verification.

Tooling transfer began with the "composable weapons" architecture Raytheon has been maturing: reusable radios, seekers, control actuation systems, and processing stacks with open interfaces. Anduril adapted that logic for Barracuda. The Polish line receives calibrated fixturing for composite airframe layup, automated winding stations for the solid-rocket booster, and a common avionics integration bench that validates the same software load whether the missile flies from a HIMARS launcher or a containerized rack. Test infrastructure is shared: the same RF chamber, vibration table, and hardware-in-the-loop simulator qualify every variant.

Workforce composition reflects the "built with, not for" model Rheinmetall and Anduril described. PGZ provides the heavy-industrial base: machinists, composite technicians, and quality engineers already cleared for NATO Secret work. Anduril embeds a small cadre of U.S. production engineers to stand up the line, then transitions to Polish hires. Specialized A&D roles in Poland command a 15–25 percent premium over comparable software positions, driven by security-clearance requirements and domain expertise.

The first articles will come off a line sized for hundreds per year, with tooling rated for the 5,000-unit annual capacity Anduril has publicly targeted for Barracuda-500 by end of 2026. Scaling means qualifying Polish sub-tier suppliers for carbon-fiber prepreg, precision-machined actuator housings, and the solid-rocket motor grain, work that feeds into the supply-chain sovereignty section.

Software Runs the Line

Anduril does not run a factory the way Lockheed or Raytheon does. Its production philosophy, branded ArsenalOS, treats manufacturing as a software problem first, a metalworking problem second. The Polish Barracuda-500M line is the first overseas deployment of that model.

At the core is Lattice, Anduril's AI-driven command-and-control platform. In the field, Lattice fuses radar, electro-optical, and electronic-warfare feeds into a single tactical picture and coordinates autonomous interceptors (Anvil, Roadrunner, Pulsar) at machine speed. On the production floor, the same architecture ingests sensor streams from automated test stations, robotic assembly cells, and supply-chain telemetry. A digital twin of each Barracuda airframe updates in real time; when a composite layup drifts out of tolerance, the twin flags it before the part reaches final integration. That loop (detect, classify, re-route) is the same kill chain Lattice runs against drones, repurposed for yield.

Traditional primes build to a drawing package, then inspect. Anduril builds to a data contract. Every workstation publishes its state to Lattice; the platform enforces the contract across the line. If a Polish sub-tier vendor ships a rocket-motor case with a marginal grain bond, the digital twin catches the anomaly at goods receipt, not at static fire. The Army's $20 billion enterprise contract, consolidating 120-plus separate procurement actions into one Lattice-backed vehicle, was awarded on exactly that premise: software-defined acquisition beats document-defined acquisition when the threat cycle moves in weeks.

The Polish line inherits this stack. ArsenalOS orchestrates tooling transfer from Anduril's U.S. Arsenal facilities (the first, a 5-million-square-foot campus, is the reference design). Automated test equipment (vibration, RF, thermal-vac) runs scripted sequences that write results straight to the airframe's digital thread. No paper travelers. No manual sign-offs. The compliance layer (ITAR, export control, Polish security clearance) is encoded as policy rules inside Lattice, so a cleared engineer in Costa Mesa can audit a torque-wrench event in Gliwice without a spreadsheet exchange.

Anduril's own hiring data reflects the shift. The company shows 2,005 open roles with a median salary band of $194k, including Head of Production, Rocket Motor Systems (McHenry, Mississippi, $225k–$397k) and Senior Director, Production Operations – Imaging (Waltham, Massachusetts, $292k–$386k). Those titles don't exist at legacy primes. They're manufacturing roles that require fluency in ROS 2, Kubernetes, and MIL-STD-810, because the line is code first, steel second.

The difference shows up in cycle time. A traditional cruise-missile line measures throughput in months per unit. Anduril targets weeks, and the Polish facility is the proof point. If the digital twin holds, the first Polish-built Barracuda-500M ships with the same data pedigree as a Costa Mesa unit. That's the sovereign promise: not just a license to build, but a license to iterate at Silicon Valley speed.

Export Control and ITAR

Moving a U.S.-designed cruise missile line onto Polish soil means building a compliance apparatus before the first rivet is driven. The Barracuda-500M falls squarely under the U.S. Munitions List (Category IV: launch vehicles, guided missiles, ballistic missiles, rockets, torpedoes, bombs, and mines), so every technical data package, manufacturing instruction, and test procedure is ITAR-controlled under 22 CFR Parts 120–130. The Directorate of Defense Trade Controls (DDTC) treats foreign production as a distinct licensing threshold: under §125.1(b), a standard technical data license (DSP-5) cannot be used for offshore manufacturing unless DDTC approves the arrangement under Part 124 procedures, meaning a Manufacturing License Agreement (MLA) or Technical Assistance Agreement (TAA) must be in place before Polish engineers touch a single controlled drawing.

That agreement creates a compliance boundary that runs through the factory floor. Technical data (defined in §120.33 as information required for design, development, production, manufacture, assembly, operation, repair, testing, maintenance, or modification, including blueprints, drawings, photographs, plans, instructions, and documentation) must be segregated, encrypted, and accessed only by authorized persons. Classified technical data adds another layer: §125.3 mandates transfers follow the DoD National Industrial Security Program Operating Manual (NISPOM), with nontransfer-and-use certificates (DSP-83) signed by the foreign government end-user. Every Polish national handling controlled data needs a security clearance recognized by both governments; every U.S. citizen on-site operates under their existing clearance but within a foreign facility subject to Polish security law.

The workforce map reflects this dual sovereignty. Anduril will need U.S.-based export compliance analysts (roles the company actively recruits for in Costa Mesa at $90k–$135k) to manage license amendments, reexport requests, and DDTC reporting. On the Polish side, PGZ must staff a mirror organization: ITAR compliance officers, empowered to control data flows, audit subcontractors, and interface with DDTC. NCAB Group Poland, a Polish PCB manufacturer registered with DDTC, shows the model: it employs named ITAR Compliance Officers, runs encrypted SFTP for controlled data transfers, and maintains rigorous licensing protocols for defense-grade boards. The Barracuda line will require equivalent roles (likely 10–15 dedicated compliance personnel across both entities), plus cleared logistics coordinators for classified component movements and a technology control plan (TCP) approved by DDTC that governs visitor access, network segmentation, and data-at-rest encryption.

Poland's NATO membership unlocks a narrow but critical exemption under §125.4(c): defense services and related unclassified technical data can flow to NATO nationals for DoD quote or bid responses without individual licenses, limited to build-to-print, build/design-to-spec, and basic research categories. But Barracuda production goes beyond that (it involves manufacturing know-how, engineering analysis, and applied research under §120.43(c), (d), (e), (i)), so the MLA/TAA route is mandatory. The compliance team's daily work will be maintaining that agreement's terms: tracking every technical data disclosure, logging foreign person access, submitting annual reports, and flagging any design changes that trigger reclassification. A single missed log entry or unauthorized screen-share can void the agreement and halt the line.

Supply Chain Sovereignty

PGZ's 50-plus subsidiaries give the Barracuda program a ready-made industrial lattice, but qualifying them for a U.S.-designed cruise missile is a different exercise than feeding legacy Polish artillery or MANPADS lines. The Bydgoszcz final-assembly site will need airframe structures, composite skins, avionics enclosures, wiring harnesses, and a solid rocket motor (SRM) supply chain that meets both NATO airworthiness and U.S. ITAR traceability.

The nearest precedent is L3Harris's qualification of Wojskowe Zakłady Elektroniczne (WZE) for PAC-3 Attitude Control Motors. That effort (specialized equipment installed at a new Zielonka facility, U.S.-based training at Camden, Arkansas, and a full Production Readiness Review) took years. WZE, a PGZ subsidiary, now produces 180 egg-sized ACMs per PAC-3 interceptor. The Barracuda line will follow a similar gate-review cadence, but across a wider component set: not just propulsion but guidance sections, control surfaces, and the Lattice-compatible mission computer.

Anduril brings its own SRM capacity (6,000-plus motors annually from its U.S. lines) and a Rheinmetall partnership announced in 2024 to develop European-sourced solid rocket motors for continental programs. That joint effort explicitly targets "sovereign suppliers and industrial partners throughout Europe." For Poland, the question is whether PGZ subsidiaries such as Mesko (explosives and propellants), Bumar-Łabędy (mechanical structures), or WZE itself can absorb the tooling and process specs for Barracuda's motor and airframe segments without creating a single-point dependency on U.S. shipments.

PGZ's own communications cite "growing integration of European components" for the Polish Armed Forces buy. What they don't yet name are the specific sub-tier vendors cleared for Barracuda drawings. The Warsaw-area composite houses that supply Airbus and Leonardo (companies like PCC Aero Structure or Polskie Zakłady Lotnicze) are plausible candidates for fuselage sections, but neither Anduril nor PGZ has published a qualified-supplier list. The avionics harness work will likely route through WZE given its PAC-3 ACM electronics experience, while warhead integration may stay with Mesko's legacy missile fuels line.

The pattern from the L3Harris/WZE playbook is clear: U.S. prime installs tooling, runs Polish engineers through U.S. production campuses, then audits the line to Manufacturing Readiness Level 8 before low-rate initial production. Anduril's software-defined approach (digital twin, automated test, Lattice integration on the factory floor) could compress that timeline if PGZ's sub-tier adopts the same data environment. But until the first Polish-built Barracuda passes a U.S. government Production Line Verification, the supply chain remains a work plan, not a proven pipeline.

Where Will the Workers Come From?

Poland's defense workforce is being rebuilt from the shop floor up. The government's "repolonization" strategy (explicit in Prime Minister Donald Tusk's visit to Mesko's new Skarżysko-Kamienna hall, where small-caliber output quintupled to 1 million rounds daily) treats skilled labor as a sovereign asset. Anduril's Barracuda-500M line with PGZ inherits that logic: the missile cannot be assembled without a workforce that did not exist at scale three years ago.

The Polish ecosystem is already calibrating. Military University of Technology (WAT), overseen by the Minister of National Defence, signed a cooperation agreement with Milrem Robotics to accelerate joint research and education in defence robotics. Kielce University of Technology inked a strategic partnership with Hanwha Aerospace and Changwon National University covering security, aviation, and space technologies. These are not pilot programs; they are the pipeline infrastructure Anduril and PGZ will tap. WAT graduates roughly 1,500 engineers annually across mechanical, electrical, and materials specializations; Kielce adds another 800 in mechatronics and automation. Neither university has a public agreement with Anduril yet, but both have the curriculum, the security clearances, and the industrial liaison offices built for exactly this transfer.

The Poland line will need a localized version of that stack: composite technicians for airframe layup, propulsion integration specialists, automated test engineers fluent in Lattice's digital twin, and a compliance cadre cleared for ITAR-controlled work on NATO soil. PGZ's existing workforce at WZL-2 (which completed three offsets projects under the WISLA Phase 1 program with Lockheed Martin) provides a nucleus, but the Barracuda's software-defined architecture demands cross-trained profiles that Polish vocational schools are only now starting to produce.

The signal for European defense tech labor markets is clear: the old split between "prime engineer" and "factory hand" is dissolving. Anduril's model (software-defined manufacturing, digital twin validation, Lattice-integrated test) requires technicians who can read telemetry, operators who can debug a firmware push, and quality engineers who treat a composite cure cycle like a CI/CD pipeline. Poland is betting that conscript-to-technician pathways, subsidized upskilling, and university-industry co-design can produce that hybrid workforce faster than Germany or France can retrain theirs. The Barracuda line is the proof point. If it works, the template replicates across the eastern flank.

Working in frontier tech? Zero G Talent tracks the openings: browse frontier tech jobs, openings at Anduril Industries, and the people building the field.