The Shipyard Problem No One Wants to Face
America can build aircraft carriers but it cannot build commercial ships anymore. Less than one percent of new merchant vessels worldwide are constructed on American soil, a statistic that would be almost comical if it weren't so strategically dangerous. A sweeping new Maritime Action Plan has finally laid out the anatomy of a crisis decades in the making, and the diagnosis is not pretty.
The Maritime Hollow
Defense Tech and Acquisition reports the raw numbers starkly: the United States operates just sixty-six total shipyards, of which only eight are active shipbuilding facilities. Strategic competitors, meanwhile, dominate the global market and construct vessels at a fraction of U.S. production costs. The piece argues that this is not merely an economic problem — it is a dependency on foreign industrial capacity that would snap shut the moment conflict erupts.
The Maritime Action Plan organizes its response around four pillars: rebuilding domestic shipbuilding capacity, reforming workforce education, protecting the maritime industrial base, and linking national security to economic security. The recommendations are familiar enough — recapitalize public shipyards, use commercial designs instead of bespoke military ones, explore public-private partnerships. But the plan also calls for something more provocative: designating zones within the U.S. exclusive economic zone, including the Great Lakes, as streamlined testing areas for autonomous maritime technologies.
A self-sustaining domestic shipbuilding sector is critical for national and economic security.
The historical echo here is unavoidable. The Merchant Marine Act of 1920 — the Jones Act — was written to ensure America could carry its own trade and defend its own coasts. Over a century later, the industrial base that act was supposed to protect has evaporated. You can write all the protectionist legislation you want, but legislation does not forge steel or train welders.
Critics might note that the plan leans heavily on technologies that do not yet exist at scale — autonomous vessels, AI-driven design optimization, additive manufacturing — as though hoping the engineering problems will resolve themselves if the policy language is confident enough. The gap between aspiration and production capacity is precisely what has undermined previous shipbuilding initiatives.
Culture Eats Policy for Breakfast
The publication's coverage shifts from ships to bureaucracy with a piece on federal acquisition reform that may be the most honest assessment of government procurement in recent memory. The barrier is not policy. It is culture.
Defense Tech and Acquisition reports that contracting officers routinely decline legally valid procurement flexibilities because they fear audits, fear leadership scrutiny, and fear being the one person who tried something different. The piece argues that other transactional authorities, commercial solutions openings, and flexible acquisition pathways are only as effective as the willingness to actually use them.
GenAI adoption inside government is happening, but slowly and at the surface level. The publication identifies the trap clearly: artificial intelligence should reimagine workflows, not automate broken legacy processes. There is little value in using machine learning to speed up a compliance checklist that should not exist in the first place.
The marketplace model for procurement — a centralized platform where agencies post needs and industry responds — is described as one of the most promising advancements in federal buying. But the publication warns against fragmentation. Industry should not have to submit fifty different solution videos to fifty different agency portals. A federated, interoperable model is the only path that scales.
Critics might note that every administration promises acquisition reform and every administration discovers that the people executing contracts have incentives diametrically opposed to taking risks. A contracting officer who tries something new and fails will face consequences. One who follows the old process and fails will not. No memo changes that math.
The Drone Gauntlet
The Pentagon is preparing to spend eleven billion dollars on more than three hundred thousand unmanned platforms through what it calls the Drone Dominance Program. Defense Tech and Acquisition reports that twenty-five vendors will compete for the first phase, but by the final round — running from August 2027 to January 2028 — as few as three vendors may receive the bulk of orders.
The structure is deliberate: each phase increases unit purchases while decreasing the number of winners. The department anticipates spending forty-five million dollars to buy one hundred fifty thousand drones in the final phase alone. The average price per unit is expected to fall as volumes spike.
But here is the tension the publication itself flags: the surge in demand for small unmanned systems, combined with a preference for American-made components, could strain the very industrial base the program is supposed to strengthen. Owen West and Travis Metz, writing for a postponed Senate Armed Services Committee hearing, noted that supply chain constraints would become visible quickly once this kind of procurement machine turns on. Once illuminated, those constraints would fall to the Department's Industrial Base Policy office and the Office of Strategic Capital.
The parallel is worth sitting with. You cannot will a domestic supply chain into existence by writing purchase orders. The same workforce shortage, the same regulatory friction, the same risk-averse culture that hollowed out shipbuilding are present in the unmanned systems sector — they have simply not been tested at volume yet.
Counter-Drone Arms Race
Running alongside the offensive drone buildout is a quieter procurement push on the defensive side. The Joint Interagency Task Force announced a two hundred million dollar agreement for the Bumblebee V2 counter-drone system — a kinetic interceptor that destroys hostile drones through direct collision. Both aircraft are rendered inoperable. It is, in essence, a kamikaze drone designed to kill other drones.
The Defense Innovation Unit is simultaneously soliciting sensor systems capable of detecting small unmanned aircraft at minimum ranges of two kilometers, with demonstrations required at Yuma Proving Ground on timelines as short as thirty days between notification and execution. The sensors must distinguish drones from ground clutter and birds, operate across multiple spectra, and be maintainable by troops in the field.
The publication does not say the quiet part out loud: the counter-drone market is chasing a threat that evolves faster than any procurement cycle can accommodate. By the time a sensor system passes demonstration, the drones it was designed to detect have already changed their data links, their flight profiles, their materials.
An AI-First Military, or an AI-First Presentation
The piece also covers the department's new AI Acceleration Strategy, framed as a declaration that the U.S. military must become an artificial intelligence-first warfighting force. Defense Tech and Acquisition reports that China and Russia are already deploying AI in surveillance, cyber operations, electronic warfare, and autonomous weapons systems — and they are not slowed by endless reviews or internal debates.
The publication argues that long pilot programs have produced glossy presentations but little battlefield impact. The new model is supposed to be simpler: build quickly, test in real conditions, deploy what works, and fix what does not. Fresh technology firms, not legacy contractors, are positioned as the primary vehicles for this transformation.
Critics might note that declaring an organization AI-first is easier than restructuring the acquisition, training, and security clearance systems that would make it possible. AI training at every rank sounds compelling until you consider that the military still struggles to get consistent IT modernization across its branches. Machine-speed decision-making requires data infrastructure that most units do not yet possess.
Augmented reality and additive manufacturing are mentioned as efficiency tools for the shipbuilding design process — promising technologies that could compress design-to-production timelines if integrated properly. But integration is precisely what has eluded defense procurement for decades. 3D printing has been "five years away from transforming military logistics" for at least fifteen years.
The future battlefield is being built right now, and the U.S. must lead it.
The Realignment Nobody Is Talking About
Buried in the coverage is a structural change with significant implications: the realignment of the Defense Security Cooperation Agency and the Defense Technology Security Administration under a single acquisition and sustainment authority. The move consolidates defense sales, arms transfers, and industrial base policy under one organization with a stated goal of improving efficiency and enabling burden-sharing with allies.
The publication presents this as a clean organizational improvement. Whether it actually reduces friction or merely moves it to a different org chart remains to be seen. Consolidation has a mixed track record in defense acquisition — sometimes it eliminates redundancy, sometimes it creates a single point of failure.
Bottom Line
The Maritime Action Plan and the Drone Dominance Program share the same blind spot: they assume that procurement volume and policy directives can resurrect industrial capacity that took generations to build and less than two to dismantle. The culture problem in federal acquisition is real, but culture does not exist in a vacuum — it is shaped by the incentives, timelines, and risk calculations that procurement systems create. Until those systems reward speed and tolerate failure, every reform memo will read like a promise the government is not structured to keep.