The U.S. defense industrial base is facing an uphill battle. A 2023 report found that 64 percent of defense firms struggled to hire skilled labor. In late 2024, the national security advisor warned that the defense industrial base “we inherited was not up to the task that we face in a new age of strategic competition.”
In April, an executive order from U.S. President Donald Trump framed the problem in similarly blunt terms: Decades of offshoring, consolidation, and underinvestment have left the United States without the capacity and workforce needed for sustained military production.
This year, Pentagon and White House officials have enacted various policies to address this fragility, with Defense Secretary Pete Hegseth going so far as to call for a renewal of America’s “Arsenal of Freedom.”
While such efforts correctly state the urgency of reform, these actions will not solve a far deeper set of problems. The United States cannot build what it needs because it does not have the workforce, the factories, or the partnerships to do so.
Policymakers talk about “surge production,” “industrial resilience,” and “allies and partners” as cure-alls for the industry’s ailments. But these terms assume that the machines, tools, robots, and networks exist and have the qualified people to operate them. Too many policymakers treat the defense industrial base as a black box: Dollars go in, and parts should come out. But no one knows which factories have modernized, which are obsolete, which are staffed, and which are silently eroding.
We contend that three structural problems lie at the heart of this dangerous fragility in the U.S. defense industrial base. First, Washington lacks basic data on modernization. Second, the skilled-labor pipeline is collapsing. And third, small yet critical suppliers are getting left behind in the push for automation.
First, the data problem. For all the political momentum behind “repairing the industrial base,” the U.S. government can’t answer the most basic question about defense production: Which factories have actually modernized? Despite years of talk about automation, robotics, additive manufacturing, and digital integration, it has surprisingly little insight into where those technologies are deployed across thousands of defense suppliers.
This is not a simple oversight. It is a structural consequence of a multidecade-long trend of outsourcing production to an opaque, and often untracked, network of private firms. Over the past two decades, periodic efforts to map the industrial base have failed to build a comprehensive database, leaving policymakers without a reliable baseline to assess the system’s health.
Unlike major aerospace companies reporting quarterly earnings or semiconductor fabs announcing capacity expansions, defense subcontractors face no requirement to disclose their operational capital investments in robotics, high-end machine tools, or industrial connectivity. While the Pentagon’s undersecretary of defense for acquisition and sustainment oversees the defense industrial base, it has little to no insight on the harsh realities facing most of these companies and subcontractors.
The legal doctrine of “privity of contract” generally restricts the Defense Department from directly auditing or demanding operational data from lower-tier suppliers, as the government’s contractual relationship typically stops at the prime contractor.
Prime contractors, or primes, often treat their supply networks as proprietary trade secrets to protect competitive advantages and frequently lack visibility into their own deep supply chains. Consequently, the government’s regulatory view stops at the prime’s front door, leaving a vast network of small subcontractors structurally opaque. Consequently, while the government may see the prime’s direct suppliers, its regulatory view effectively fades to black beyond the first tier, leaving the vast network of lower-level subcontractors structurally opaque.
This blindness creates a major planning error. Defense planners routinely overestimate surge capacity because they assume the industrial base can scale linearly with funding. If a plant produces 30 Patriot interceptors per month, they assume it can produce 60 with double the money. But doubling output requires far more than doubling dollars. It requires knowing what the reality looks like on the ground. Are machines aging out? Does the facility have the skilled labor it needs to expand? Without this critical information, surge projections become little more than optimistic extrapolations.
The majority of defense programs already face serious labor shortages. Take the submarine industry, which needs about 100,000 new specialized workers, from precision machinists to pipe fitters, over the next decade to meet the Navy’s commitments for building Virginia-class and Columbia-class submarines. Most will need security clearances.
No other industrial sector has such a fraught combination of skill intensity and clearance requirements, and no education pipeline is producing workers at anywhere close to the needed scale. Achieving mastery in critical defense trades demands sustained apprenticeship and five to 10 years of experience—a developmental horizon rarely appreciated in short-term policy planning.
The same story plays out across the missile, space, and munitions sectors. The defense industrial base desperately needs machinists to run 5-axis computer numerical control (CNC) mills, which are essentially computer-controlled metal-cutting robots. It needs welders capable of joining exotic alloys; automation technicians who can keep robotic welding cells online; and controls engineers who can integrate sensors, digital twins, and industrial networks. These occupations face an aging workforce, declining apprenticeship rates, and stiff competition from commercial sectors with better pay and no security clearance.
Policymakers often assume that installing robots will compensate for labor shortages. Yet automation actually increases the need for higher-skilled workers: robot programmers, maintenance technicians, industrial-network specialists, and cybersecurity operational-technology personnel. A single robotic cell may eliminate two low-skill roles but require five advanced technicians to keep it running under the Pentagon’s standards. While a single robotic cell may eliminate repetitive manual roles, it creates a net new demand for advanced technicians to program, maintain, and secure the system—roles that are far harder to fill.
The most overlooked problem may be the thousands of small factories that form the backbone of the U.S. defense industrial base. They are not household names because they usually employ fewer than 50 people and rely on second- or third-generation owners. But their products are indispensable for critical technologies such as submarine propulsion systems, Patriot missile canisters, or housings for Javelins’ seekers.
They are also the places where the industrial base is its most vulnerable, as modernization has largely passed them by. A prime might show Congress a futuristic digital production line with robotic arms, laser trackers, and advanced metrology. At the same time, one of its tier-three suppliers might be machining missile components on a 1987 Haas mill with no digital controls, no cybersecurity, no automation, and no ability to surge production. Most of these shops are running legacy equipment because they do not have the sort of profit margins or shareholders to replace expensive equipment. Instead, they operate in a contracting environment where Pentagon awards are often unpredictable, payments are slow, and margins are thin.
This explains why small suppliers are exiting the market faster than they can be replaced. A 2022 report from the Defense Department revealed that the number of small businesses in the defense sector shrank by over 40 percent in the last decade, despite increased spending. According to a companion report, this attrition was most acute in the specific sub-tiers required for surge capacity. The Pentagon explicitly designated castings and forgings and kinetic capabilities as among the top-priority sectors for immediate federal intervention.
Further compounding the problem, prime contractors rarely know the full state of their own subcontractor networks; the government knows even less. In a world of fragile supply chains and peer competition, the industrial base is only as strong as its most brittle tier-three supplier. The United States can spend billions on new programs, but if a single small shop in Ohio or Arkansas goes out of business, an entire weapons line can grind to a halt.
The final barrier is institutional. If the United States cannot modernize its industrial base alone, the obvious solution would be to distribute production among its allies. In theory, the United States has never had a stronger network. Australia, Japan, South Korea, and most NATO members possess world-class machine-tool sectors, advanced robotics clusters, precision-casting facilities, and highly automated microelectronics plants.
South Korea produces some of the most modern armored vehicles and artillery systems in the world. Japan and Germany lead in ultraprecision machining. The United Kingdom and Australia are already aligned under the AUKUS partnership to develop next-generation submarine technology and autonomous systems. In many of these sectors, these allied nations are not just partners; they are the global standard that the U.S. is now struggling to match.
On paper, this reads like the foundation for the new Arsenal of Freedom that policymakers envision. In practice, however, these partnerships collide with a bureaucratic problem. American companies cannot share technical data quickly or easily enough to make coproduction feasible. The same system designed to safeguard sensitive defense technology—such as export controls, classification barriers, and strict disclosure rules—also makes it difficult to build seamless production networks even with America’s most trusted allies.
This obstacle is not merely legal—it is entrenched in the defense industry’s culture. The Department of Defense’s Cold War-era “Not Releasable to Foreign Nationals” policy has become one of the most significant challenges to the very allied coproduction and technological integration that modern defense strategy now seeks to achieve. It is a problem nearly every AUKUS and National Technology and Industrial Base (NTIB) working group faces when collaborating: excessive disclosure rules that delay or prevent allied engineers from accessing the designs required to build or test components.
The United States does not need another new alliance to solve this problem. It needs to allow its existing allies to help.
The uncomfortable truth is that policymakers cannot fix the defense industrial base until they peer inside the black box and make sense of it as a real and complex system of machines, workers, suppliers, and allied partners.
First, the Pentagon should build an “advanced manufacturing readiness” baseline. It should require a standardized and anonymous reporting framework for defense suppliers to disclose the age, capability, and digital maturity of key production assets. Without this visibility, every effort to scale production is guesswork. A national map of modernization, tracked and updated across the supplier network, would enable the Pentagon to target investments intelligently.
Second, invest in workers, not just workshops. America needs a deliberate pipeline for cleared machinists, welders, avionics technicians, automation specialists, and cyber workers. That means expanding programs and opportunities in high schools, community colleges, union apprenticeships, and technical credentialing for high-skill trades, as well as fast-tracking security clearance pathways for industrial workers.
Third, modernize the small suppliers that carry the real load. The federal programs that fund industrial-base upgrades and targeted modernization initiatives must prioritize tier-two and tier-three firms for new equipment, digital integration, cybersecurity, and automation support. Grants and cost-sharing mechanisms can help small shops acquire high-end machine tools, robotic cells, or industrial networking. These are the kinds of investments that prime contractors can afford but the small businesses on which they depend cannot.
Finally, the United States needs to unlock allied coproduction by reforming disclosure rules. Partnerships like AUKUS and NTIB can only function if allied engineers can access the technical data needed to build, test, and integrate components. Reforming export controls for pre-approved partners, enabling digital technical-data environments, and delegating more disclosure authority within the Pentagon would turn allied goodwill into actual production.
American industrial strength must be more than a slogan. It is the sum of machines, workers, suppliers, and partners. The sooner Washington opens this black box, the sooner it can rebuild the arsenal it believes it already has.
