The Arsenal of Democracy Can’t Be Backordered

By Michael Baumgartner

Wednesday, December 17, 2025

 

Last week, Congress quietly did something on a broad bipartisan basis that could matter a great deal in the next major war. It passed the National Defense Authorization Act (NDAA) for Fiscal Year 2026, a bill that starts to redress America’s broken arms-sales system, speeding weapons to frontline allies who are already staring down China, Russia, and Iran.

 

The new NDAA creates a fast lane for close partners like Australia and moves more of our best equipment out of the slowest government-only channels and into faster commercial routes. Specifically, it codifies key reforms from President Trump’s April 9 Executive Order on foreign defense sales — cutting red tape, improving transparency, and ensuring more defense items can move through faster pathways to trusted allies. It’s a welcome step — but rather than taking a victory lap, we should view this as just a starting point. We need to push harder to get to the 21st-century defense system we need.

 

America didn’t win World War II because we wrote better regulations. We won because we turned places like Detroit, Pittsburgh, and my hometown of Spokane, Wash., into an arsenal of democracy that outbuilt and out-invented our enemies at industrial scale. “Freedom’s Forge” is not nostalgia; it’s the standard we must meet in an era of AI, drones, and hypersonics.

 

The decisive technologies of the next war won’t be found in steel mills and shipyards alone. They’ll be chips and AI models, autonomous systems and drone swarms, space-based and cyber tools, and synthetic biology and biodefense. And the “arsenal” won’t be a few smokestack plants in one country. It will be a web of cloud data centers, AI labs, chip fabrication plants, and production lines that run from Spokane to Seoul and from Huntsville to Helsinki.

 

Our adversaries understand this. China is openly investing in AI-enabled weapons, drone swarms, and anti-satellite systems. Russia is improvising cheap, semi-autonomous kamikaze drones on the battlefield in Ukraine. Iran and its proxies are already experimenting with low-cost, high-volume unmanned systems to probe U.S. forces and our allies.

 

Meanwhile, too much of the American defense system is still running on a late-Cold War operating theory: that America will always enjoy a decisive technological edge, that we can hoard the crown jewels behind thick export controls, and that a small domestic cartel of defense contractors can scale on demand to arm us and our allies. None of that is guaranteed anymore.

 

Some in Congress recognize the risks. I serve on the House Foreign Affairs Committee’s Arms Sales Task Force, which has exposed how our own bureaucracy turns foreign military sales into a strategic liability — leaving allies like Taiwan waiting years for already-approved weapons while Beijing arms at speed. The reforms we just passed matter. They point in the right direction. But they’re still just a first turn of the wrench on a system built for a different century.

 

We need something bigger: a Freedom’s Forge for the 21st century. This will require four big shifts.

 

First, we should build on AUKUS, the trilateral partnership between Australia, the United Kingdom, and the United States, and establish fast lanes for more allies — especially those on the front lines against China, Russia, and Iran — with clear, enforceable timelines for licensing and delivery. At the same time, we need to abandon the mindset that we can wall off entire fields and instead put a hard brake on the specific enabling technologies that matter most to our enemies: advanced AI chips, the giant data centers that run them, and advanced biomanufacturing tools. The tools to design a drone swarm or engineered pathogen that threatens Americans should not be licensed by the U.S. or allied governments in the first place.

 

Second, we must rebuild industrial capacity in certain strategic sectors. Maintaining an edge in munitions, drones, air and missile defense, space, energy storage, critical minerals, and microelectronics requires multiyear procurement, allied co-production, targeted investment in surge lines, and real market access for new, innovative firms.

 

Third, we ought to treat AI and synthetic biology like nuclear physics — but with a very different playbook. Our knowledge of nuclear technology is locked in a few guarded government labs — but AI and synthetic biology tools live in global commercial markets and thousands of labs, so we can’t hoard them in a vault. Rather, the goal should be to set enforceable rules — with the buy-in of our allies, since the supply-chain-critical chokepoints are not all American — for how they’re built, distributed, and used. That means clear national standards, tighter control over the most sensitive inputs like extreme ultraviolet lithography, high bandwidth memory, and advanced DNA synthesis. The approach we take here must be aimed at sustaining a decisive, American-led, allied edge rather than assuming one is guaranteed.

 

And lastly, we should reinvigorate research universities and national labs. This partnership between science and government is critical terrain in the competition to define and guide the future: These research centers train the scientists and engineers, run the labs, and spin out the companies that will determine who leads in AI, chips, and bio. We should fund the next generation of AI, quantum, materials, and bio research. Universities need the freedom to innovate, but we cannot continue to allow China and other nations to steal intellectual property and influence governance through donations and front organizations. And a serious institution will not allow antisemitism and woke ideology to dominate its academic and research culture. Government should fortify the labs, not abandon them.

 

We can either build the future arsenal of democracy on our terms and those of our allies — or watch our enemies build theirs and write the rules for us.