A clash between the Pentagon and the White House over a proposed ban on Chinese rare earth magnets has exposed the foundational contradiction at the heart of US critical mineral policy: the national security imperative to decouple from China is colliding with the defense industrial base's near-total reliance on Chinese-sourced permanent magnets. With a statutory deadline arriving January 1, 2027 and Beijing's export control suspension expiring just weeks before, the window to resolve that contradiction is closing fast.
Introduction
In Johnston County, North Carolina, a startup called Vulcan Elements is preparing to build what its founders describe as the largest rare earth magnet factory outside China. The facility does not yet exist. The company is three years old. And yet by January 2026, it carried a reported valuation of nearly two billion dollars, a tenfold increase from the two hundred million dollars at which investors priced it the previous August, when a venture fund co-managed by Donald Trump Jr. took an undisclosed stake in its sixty-five million dollar Series A round. Three months after that investment, the federal government committed six hundred and seventy million dollars to the company, including a six hundred and twenty million dollar conditional loan from the Pentagon's Office of Strategic Capital, pushed through in a matter of weeks at the direction of Peter Navarro, a senior White House counsellor and personal friend of Trump Jr.
The Vulcan story might read as a straightforward corruption narrative, the kind of thing that produces Senate letters and House subpoenas, which it has. But it is also something more uncomfortable: a symptom of the extraordinary pressure bearing down on Washington as it tries, belatedly and chaotically, to build a domestic rare earth supply chain before a series of overlapping deadlines converge into something that looks very much like a crisis.
In the week of May 19 to 24, Bloomberg reported that Pentagon officials had raised serious doubts about the viability of an eighty million dollar loan to ReElement Technologies, another critical minerals company in the broader fourteen-hundred-million-dollar initiative, triggering a clash with the White House over the pace and rigour of the vetting process. Days later, the Financial Times reported that a coalition of defence prime contractors was lobbying the Trump administration to delay a statutory ban on Chinese-sourced rare earth magnets scheduled to take full effect on January 1, 2027, citing the absence of any viable domestic or allied alternative at scale. And hovering above all of it: a Chinese export control suspension that expires on November 10, 2026, one month before the ban takes effect, and which Beijing can reinstate with thirty days' notice.
The Deadline Architecture
To understand why the Pentagon and the White House are fighting over an eighty million dollar loan to a company that has not yet produced rare earth oxides at commercial scale, it helps to understand the regulatory vise that is tightening around the entire US defence industrial base.
Under 10 U.S.C. Section 4872 and the Defence Federal Acquisition Regulation Supplement provision 252.225-7052, effective January 1, 2027, US defence contractors are prohibited from delivering any covered material containing neodymium-iron-boron or samarium-cobalt magnets mined, refined, separated, melted, or produced in China, Russia, Iran, or North Korea. The rule does not simply restrict finished magnets purchased from Chinese suppliers. Its traceability mandate reaches all the way back through the supply chain to the mine itself. A rare earth element extracted in Montana or Saskatchewan but processed through a Chinese refinery does not satisfy the non-Chinese-origin requirement. The chain of custody must be clean from extraction to finished component, with auditable documentation at every stage.
The legislative origins of this prohibition trace back to the first Trump administration. Congress introduced the framework in 2018, making this a genuinely bipartisan creation, one of the few areas of American industrial policy where the two parties have maintained something resembling consensus. Its strategic logic is straightforward: no weapons system ought to depend on materials controlled by an adversary. Its practical challenge is equally straightforward, and has been visible for years to anyone paying attention: the United States and its allies still lack the downstream manufacturing capacity to comply with it.
Abigail Hunter, who directs the Center for Critical Minerals Strategy at SAFE, Securing America's Future Energy, put the tension plainly when the Financial Times reported on the lobbying campaign in mid-May. The ban, she argued, would allow the United States, not an adversary, to define when, what, and how it sources materials for its defence industrial base. But she also acknowledged the operational reality: "Now that we are facing ongoing conflict and the urgent need to replenish our weapons systems, we may need flexibility in execution, but we absolutely cannot dilute the original bipartisan policy intent." The phrase "flexibility in execution" is doing a great deal of work in that sentence.
The Weight of Dependency
The statistics describing America's reliance on Chinese rare earth magnets have been cited so often that they have begun to lose their power to shock. So it is worth pausing on what they actually mean in physical terms, in the weight of materials embedded in specific platforms that the US military depends on.
A single F-35 fighter jet contains more than nine hundred pounds of rare earth materials. A Virginia-class submarine requires over ninety-two hundred pounds. An Arleigh Burke-class destroyer carries nearly six thousand pounds of rare earth content. Across the US defence industrial base, at least eighty thousand components in nineteen hundred weapons systems depend on Chinese-sourced rare earths. Those components include not just drone motors, where the dependency is widely discussed, but guidance systems, radar, sensors, and virtually every advanced platform the Pentagon fields. Lockheed Martin, Northrop Grumman, and RTX will all need to trace and certify their magnet supply chains before the January 2027 deadline.
The drone dimension deserves particular attention. The Pentagon recently placed the largest drone order in American history: thirty thousand one-way attack platforms, with plans to scale past three hundred thousand by early 2028. Every one of those drones runs on a rare earth magnet. Scaling from thirty thousand to three hundred thousand represents a tenfold increase in magnet requirements within a supply chain that currently has no credible non-Chinese alternative at meaningful production volumes. The defence budget for 2026 allocated thirteen point six billion dollars for autonomous systems. The rare earth supply to sustain that ambition has not been secured.
The specific elements that matter most for military applications are dysprosium and terbium, the heavy rare earths added to base neodymium-iron-boron alloys to prevent the magnets from losing their magnetic coercivity at the elevated operating temperatures of combat drone motors and missile guidance systems. These are also the elements where China's dominance is most complete. As of April 2025, China accounted for approximately ninety-nine percent of global heavy rare earth processing. The city of Wuxi alone is home to the world's only facility capable of refining dysprosium at scale. S&P Global has tracked dysprosium and terbium exports running roughly fifty percent below their pre-restriction baseline since China's April 2025 controls took effect, with no confirmed timetable for normalisation.
For countries that need these materials outside China, the price premium is severe. In some European markets, rare earth prices reached six times their Chinese domestic equivalent in 2025. Neha Mukherjee, research manager at Benchmark Mineral Intelligence, has noted that "the heavier the rare earths, the higher the cost of refining," while China can do so "at a very low cost and very efficiently." That efficiency is not accidental. It is the product of decades of deliberate industrial policy, beginning with government-directed investment in processing infrastructure in the 1990s and 2000s, at a time when Western producers were exiting the market due to cost pressures and environmental compliance burdens. In 1992, Deng Xiaoping declared that the Middle East had oil and China had rare earths. His government backed the sector with cheap loans, loose environmental rules, and direct investment. Three decades later, China controls roughly ninety-four percent of global permanent magnet production, up from around fifty percent two decades ago.
The Suspension That Is Not a Settlement
Complicating the picture is a diplomatic arrangement that the Trump administration has repeatedly characterised as a resolution to the rare earth problem, but which industry compliance experts and trade lawyers describe in considerably more cautious terms.
Following the October 30, 2025 meeting between Presidents Trump and Xi in Busan, South Korea, China's Ministry of Commerce issued Announcement No. 70 on November 7, pausing a suite of expanded export controls, known as MOFCOM Announcements 55 through 62, until November 10, 2026. The suspended package included the most aggressive instrument China had ever deployed in this domain: a so-called 0.1 percent rule, asserting Chinese export control jurisdiction extraterritorially over any product anywhere in the world that contains more than 0.1 percent Chinese-origin rare earth by value. Under this framework, a motor manufactured in Germany using Chinese neodymium could require a Chinese export licence to ship between two non-Chinese countries. MOFCOM Notice 61 broke the conventional model of export control law by applying jurisdiction to products manufactured entirely outside China's borders, by non-Chinese companies, in non-Chinese facilities.
The suspension of that framework has provided some breathing room. But it has not suspended China's April 2025 controls, which remain fully active and cover seven heavy rare earth elements including dysprosium, terbium, yttrium, and scandium. Nor has it created anything resembling predictable market access. As one industry insider described the situation to me, "MOFCOM didn't open the gate; it just oiled the hinges." Export licence applications still require extensive documentation, can run hundreds of pages, and must be justified to multiple agencies. Of roughly two thousand applications submitted by EU companies in 2025, just over half had been approved by November, according to Reuters.
The legal framework for the suspended controls has not been dismantled. It has been paused. November 10, 2026 is a confirmed statutory expiration date. If MOFCOM does not announce a further suspension before that date, the full extraterritorial architecture reimpose simultaneously on every product in the affected categories. Given that global supply chains have not materially reduced their dependence on Chinese-origin rare earth materials in the six months since the suspension began, the reimposition of those controls would create immediate disruption across automotive, defence, medical device, and industrial equipment supply chains worldwide. The IEA's April 2026 report places the full-implementation risk at six point five trillion dollars annually in at-risk economic output for countries outside China.
As I noted in my analysis of the Busan countdown in May, the diplomatic truce has provided a pause for recalibration, not reconciliation. Beijing's April 2025 controls on heavy rare earths continue to suppress export volumes. The suspension is real, but so is the cliff edge it is deferring. Industry compliance experts are now urging companies to treat the current window as an opportunity to apply for general licences while approval conditions remain more favourable, ahead of potential regulatory tightening in late 2026. The risk of waiting is that the bureaucratic backlog alone could disrupt shipments even if Beijing does not formally reimpose the suspended measures.
Investment at Trump Speed
Against that backdrop, the disputes over ReElement and Vulcan Elements look less like isolated controversies and more like a stress test of whether Washington can build a domestic rare earth supply chain at all, let alone at the pace the situation demands.
ReElement Technologies represents an intellectually interesting model. Its proposed approach focuses on secondary refining, the extraction of rare earth oxides from electronic waste and end-of-life magnets rather than from virgin ore. This circular economy approach targets a segment of the rare earth supply chain that is genuinely underinvested domestically. The problem, as Pentagon officials reviewing the deal have noted, is that as of late 2025, the company had not yet produced oxides at commercial scale, and its long-term revenue forecasts have drawn internal scrutiny. The eighty million dollar conditional loan announced in November has not been cancelled, and no funds have been disbursed. But the fact that Pentagon vetting officials are raising doubts about the deal's viability has put them on a collision course with a White House that, by its own account, is committed to moving at what it calls "Trump Speed."
The Vulcan case is more troubling in its procedural particulars. ProPublica's investigation, published on May 28, documented that the request to loan hundreds of millions of dollars to a firm connected to Trump Jr. was made personally by Peter Navarro, making it the only deal among dozens under Pentagon consideration at the time to have been initiated by a top aide to the president. Pentagon staff were directed to close the deal in weeks rather than the months the Office of Strategic Capital typically requires for due diligence. "The call came from the White House: We have to get this done," one defence official told ProPublica. The Pentagon has denied that any company received preferential treatment. Trump Jr.'s spokesperson said he has no knowledge of how the deal came together. The White House said the administration is working in the best interest of the American people.
The Senate demands for answers from Defence Secretary Pete Hegseth, sent in January by Senators Elizabeth Warren, Andy Kim, and Richard Blumenthal, have gone unanswered. Democratic House members moved to subpoena Trump Jr. to testify; the Republican majority refused. Whether the sequence of events proves anything unlawful or represents mere coincidence, it is, as one observer noted, exactly the kind of arrangement that erodes public confidence in an industrial policy that depends on public trust to function.
Beyond the controversy, the harder question is whether any of this investment is building fast enough to matter. The Department of Defense has committed four hundred million dollars to MP Materials, the only currently integrated US mine-to-magnet producer, which produced a record twenty-five hundred and ninety-nine metric tons of neodymium-praseodymium oxide in 2025, a one hundred and one percent increase year over year. The company remains unprofitable and is at the beginning of its magnet manufacturing ramp-up. Lynas, the leading non-Chinese rare earth producer, produced a combined eight tons of dysprosium and terbium in the first quarter of 2026. Global appetite for those two elements alone runs into the thousands of tons annually. Energy Fuels Inc. opened a pilot facility in Utah last year and has produced roughly forty kilograms of dysprosium so far. The Saskatchewan Research Council facility in Saskatoon, backed by REalloys and supported by the US Defence Logistics Agency, is scheduled for commissioning by December 2026 and operation in 2027, targeting five hundred and twenty-five tonnes of neodymium-praseodymium, thirty tonnes of dysprosium, and fifteen tonnes of terbium annually under REalloys' preferred supply agreements.
Benchmark Mineral Intelligence data shows that investments totalling six point three billion dollars were announced in 2025 for projects outside China, with more than sixty percent coming from US government sources, and a further two point eight billion dollars followed in the first quarter of 2026. But the consultancy's own projections, alongside those of McKinsey and CRU Group, suggest that by 2035, non-Chinese supply will cover only around half of mining requirements, a quarter of refining needs, and less than a fifth of magnet demand. For dysprosium and terbium specifically, countries outside China will still meet less than a fifth of demand a decade from now. Michel Van Hoey, a senior McKinsey partner on metals and mining, has offered the assessment that "meaningful diversification will take longer than many anticipate." It is a sentence that carries particular weight when the statutory deadline is less than nineteen months away.
Tim Johnston, co-founder of REalloys, has identified where the real bottleneck lies, and it is not where most of the political attention has been focused: "Most people talk about mining. The bottleneck isn't the mine. It's the step where oxide becomes metal and metal becomes alloy. Metallization is the least developed part of the value chain outside China." The DFARS traceability mandate requires a clean chain of custody from extraction to finished component. But the facilities capable of performing the intermediate metallisation steps outside Chinese control simply do not exist at the required scale. The US has, as Gracelin Baskaran of CSIS has noted, deployed the entire industrial policy toolkit, equity, concessional financing, public procurement. What it cannot manufacture is time.
The Convergence
The timeline that emerges from all of this has a particular quality of inevitability. China's April 2025 controls on seven heavy rare earth elements remain fully active and have suppressed dysprosium and terbium export volumes by roughly fifty percent below pre-restriction baselines. The suspended October 2025 controls, including the extraterritorial 0.1 percent rule, expire on November 10, 2026, with the legal infrastructure in place to reimpose them at thirty days' notice. One month later, on January 1, 2027, the DFARS ban takes full effect, prohibiting defence contractors from delivering any covered weapon system incorporating Chinese-origin magnets. In between those two dates, the US defence industrial base is expected to have somehow certified clean supply chains for eighty thousand components across nineteen hundred weapons systems.
Defence contractors lobbying for a delay are not arguing in bad faith. They are describing a real arithmetic problem: the US currently imports approximately ten thousand tonnes of rare earth magnets annually, and any new entrant to the qualified supply base faces a multi-year certification pathway that cannot be compressed to match political timelines. The lobbying campaign reflects not an unwillingness to comply but an acknowledgement that compliance infrastructure does not yet exist.
Beijing's management of this situation has been, by contrast, notably deliberate. China's export controls, as defence analysts have observed, appear calibrated to maximise leverage over this exact transition period, demonstrating the consequences of dependency before Western alternatives reach commercial scale. By requiring case-by-case licence approvals rather than imposing a formal prohibition, MOFCOM can selectively delay, approve, or deny individual shipments without technically violating formal trade commitments. The resulting uncertainty forces international buyers to hold larger inventory buffers, pay premiums for alternative sources, and absorb elevated supply chain costs. This regulatory ambiguity is arguably more economically disruptive than a hard export ban would be, and it is, as one trade lawyer has noted, one of Beijing's preferred strategic instruments. Given that rare earths represent one of China's most consequential levers over high-impact industries including defence, it is very unlikely that Beijing will relinquish that influence voluntarily. What it may do is calibrate the pressure with enough precision to maintain the cooperative tone of post-Busan diplomacy while preserving maximum future leverage.
Conclusion
Back in Johnston County, North Carolina, the site where Vulcan Elements plans to build its magnet factory remains, for now, an aspiration mapped onto a business plan. Whether the six hundred and seventy million dollars in federal commitments will produce a functioning facility, and whether that facility will produce magnets that satisfy DFARS traceability requirements in time to matter, remains genuinely uncertain. The company's rapid valuation trajectory, from two hundred million dollars in August 2025 to nearly two billion dollars by January 2026, reflects the extraordinary premium the market is placing on anyone who can credibly claim to solve this problem.
That premium also reflects how narrow the credible solution space actually is. The ReElement dispute, the Vulcan controversy, the contractor lobbying campaign, the November expiration, the January deadline: these are not separate stories. They are different facets of a single, unresolved crisis of strategic dependency that the United States has been narrating for years without resolving.
The internal friction that erupted in the week of May 19 to 24 between the Pentagon's due diligence instinct and the White House's urgency to move at what it calls Trump Speed is a microcosm of the larger problem. Speed without rigour produces Vulcan. Rigour without speed produces a supply chain that is still not built when the clock runs out. Neither response, on its own, is adequate to the scale of what China has constructed over thirty years of deliberate, patient, governmentally coordinated industrial policy.
"The Middle East has oil; China has rare earths," Deng Xiaoping said in 1992. Thirty-four years later, the United States is discovering, with some urgency, that he was not speaking metaphorically. He was describing a strategic reality that would take generations to construct and cannot be undone in months. Washington has perhaps begun to understand the dimensions of what it is up against. Whether it has begun to act with the sustained discipline the problem actually requires is a different question, and the answer, as the clock ticks toward November, remains far from clear.