Full Research Report: US critical minerals and rare earth strategy: breaking dependence on China after

Prepared: April 2026 | Classification: Open Source Intelligence Synthesis | Timeframe: April 2025 – April 2026

The United States is engaged in the most consequential resource competition since the oil shocks of the 1970s. China's weaponization of rare earth elements and critical minerals in 2025 — beginning with MOFCOM Announcement No. 18 on April 4, 2025, imposing export licensing on seven medium and heavy rare earths — has transformed a latent structural vulnerability into an operational crisis. The immediate effects were severe: China's total exports of rare earth magnets fell 74% year-on-year in May 2025 to 1.2 million kilograms, with shipments to the United States plummeting 93.3% to just 46,000 kilograms (single-source: e2open) e2open. Ford Motor Company suspended Explorer SUV production for one week at its Chicago assembly plant due to magnet shortages Persistence Market Research. European rare earth prices reached up to six times Chinese domestic prices IEA.

The organizing analytical concept of this report is "licensed scarcity" — China's most operationally dangerous form of leverage, in which Beijing selectively throttles export license approvals to particular firms, end-uses, or countries without triggering a formal embargo, remaining formally within WTO-compatible rules while degrading supply in ways that are difficult to attribute, challenge, or counter. Licensed scarcity is not a theoretical risk: it was demonstrated concretely during the May 2025 Swiss talks, when U.S. manufacturers began shutting down production as China delayed export license issuance despite not formally abandoning the agreed truce. This mechanism — not a total cutoff, not a formal embargo, but calibrated administrative friction — is the primary instrument through which China will exercise leverage during the 2026–2028 vulnerability window, and it is the central threat that U.S. strategy must be designed to defeat.

Bottom line: The U.S. response — anchored by the $550 million DoD-MP Materials partnership, Presidential Proclamation 11001 under Section 232, the $12 billion Project Vault stockpile, and the 54-nation FORGE coalition — represents the most aggressive critical minerals mobilization in American history. However, the structural chokepoints — China processes approximately 85% of global rare earth output, manufactures nearly 90% of rare earth magnets, and holds a near-monopoly on heavy rare earth separation IEA — cannot be resolved within the one-year truce window negotiated at the October 2025 Xi-Trump summit. The period from 2026 to 2028 is one of maximum vulnerability: domestic capacity is under construction but not yet operational, allied supply chains are forming but not yet binding, and China retains the ability to escalate through licensed scarcity at any moment of its choosing. The realistic strategic endpoint is not independence but managed, diversified interdependence, with a "China-light" ecosystem capable of sustaining defense and critical commercial applications through prolonged restrictions.

Confidence level: High confidence in the assessment of structural vulnerability and the timeline of U.S. capacity buildout. Medium confidence in the durability of the October 2025 truce and the operationalization of FORGE. Low confidence in projections of Chinese internal compliance dynamics and the commercial viability of domestic processing at scale without sustained government price support.

China's position is the product of four decades of deliberate industrial policy, not market accident. From 2000 to 2021, Beijing channeled almost $57 billion into mineral extraction and refining across nearly 20 countries in Africa, Latin America, and Asia Chatham House. China holds approximately 222,000 of the world's 470,000 rare earth patents Asia Times. The IEA's Global Critical Minerals Outlook 2025 found that for 19 of 20 strategic minerals, China is the leading refiner, with an average market share of approximately 70% IEA. China's share of sintered permanent magnet production has risen from approximately 50% two decades ago to 94% today IEA.

Beijing's 2025–2030 rare earth policy consolidation aims at "full-chain control" (全链条管控) across exploration, mining, separation, alloying, magnet manufacturing, and export governance, backed by state-owned champions — principally China Rare Earth Group and China Northern Rare Earth Group High-Tech Jamestown Foundation. Under the current Five-Year Plan, Chinese firms acquiring foreign mines must ship ore back to China for processing, ensuring the world remains dependent on Chinese factories for finished products like permanent magnets. China's 2024 mining quota stood at 270,000 metric tons of rare earth ore, with smelting and separation at 254,000 tons. For 2025, China deviated from tradition by delaying the usual Q1 public announcement and instructing recipients to keep quota details confidential for "security purposes." [Unverified intelligence — requires confirmation: this claim is cited in multiple industry analyses but no official MIIT document is publicly available confirming the confidentiality instruction; it should not be weighted in operational planning without primary source verification.]

China's strategic calculus, as analyzed by Resources for the Future, is that its export controls are not designed for long-term withdrawal of supply but function as a calculated play in a repeated game — asserting dominance and signaling resolve while leaving room for reversibility to avoid triggering irreversible shifts in global trade dynamics RFF. This mirrors the 2010 playbook against Japan over the Senkaku Islands dispute, where reductions in exports were minimal in practice but the political signal was unmistakable. The 2010 episode is instructive in a second respect: Japan's subsequent decade-long diversification effort reduced China's share of its REE imports from 85% to 58% by 2020 — demonstrating both that diversification is achievable and that it requires sustained commitment over a timeframe measured in years, not months. The current U.S. challenge is structurally harder than Japan's 2010 challenge, because China has since extended its dominance from mining into midstream processing and magnet manufacturing in ways that were not yet fully consolidated a decade ago.

However, some analysts at the Jamestown Foundation and East Asia Forum note internal tensions: Chinese firms focused on global market share and profitability may resist mandates forcing domestic processing when offshore facilities are more efficient, and local governments reliant on export revenues push back against tight controls Jamestown Foundation. This is an obscure variable with limited open-source verification, but it suggests Beijing's policy is not monolithic. It should not be weighted heavily in planning, but it does create potential fissures that patient diplomacy and targeted commercial engagement could exploit at the margins.

China has also amended its Foreign Trade Law to expand countermeasure authorities outside WTO mechanisms Covington & Burling, signaling a preference for direct state-to-state confrontation over legal arbitration. No WTO disputes have been filed challenging China's 2024–2025 export restrictions on gallium, germanium, antimony, or rare earths — a notable absence given the successful 2014 WTO ruling (DS431) that forced China to eliminate export quotas and duties on rare earths, tungsten, and molybdenum by January 1, 2015. This absence is analytically significant and deserves direct examination. The Appellate Body's dysfunction since 2019 is one reason: any challenge would proceed only through first-instance panels, with no binding appellate review. But the strategic calculation runs deeper — filing a WTO challenge signals that the complainant regards the rules-based system as the appropriate venue, which implicitly constrains the complainant's own freedom to impose retaliatory measures outside WTO rules. It also takes years to adjudicate, during which the supply disruption continues. China's invocation of the Article XXI national security exception — which WTO panels have interpreted with increasing deference since the 2019 Russia-Ukraine transit dispute ruling — further reduces the probability of a successful outcome. The net effect is that the rules-based trading system has been effectively neutralized as a check on mineral weaponization, and the U.S. and its allies are operating in a mercantilist environment that requires mercantilist tools.

The U.S. entered 2025 with a structural deficit that Congressional advocates have described as having "funded the factory, forgot the feedstock" — a reference to the CHIPS Act's $280 billion investment in semiconductor fabrication without securing upstream mineral supply Prosperous America. As of 2024, the United States was 100% net-import reliant for 12 critical minerals and at least 50% reliant for 29 additional critical minerals Federal Register — Proclamation 11001.

The Trump administration's response has been multi-pronged and unprecedented in scale. The key policy instruments, in chronological order:

Executive Order 14272 (April 2025): Initiated the Commerce Department Section 232 investigation into whether imports of processed critical minerals and derivative products (PCMDPs) threaten national security White House.

DoD-MP Materials Partnership (July 2025): The Department of Defense agreed to purchase $400 million of newly created preferred stock convertible into common shares and provide a $150 million loan to expand heavy rare earth separation at Mountain Pass MP Materials. DoD committed to long-term purchases of up to 7,000 tonnes per year for ten years, with a price floor of $110 per kilogram for NdPr products (single-source: Columbia SIPA/CGEP) — approximately $50/kg above spot price Columbia SIPA/CGEP. Commercial financing includes $1 billion in debt from JPMorgan Chase and Goldman Sachs for the 10X facility (single-source: Columbia SIPA/CGEP) Columbia SIPA/CGEP.

Presidential Proclamation 11001 (January 14, 2026): Issued under Section 232 of the Trade Expansion Act of 1962, directing the Secretary of Commerce and U.S. Trade Representative to negotiate agreements with other countries, considering price floors and other trade-restricting measures. If agreements are not reached within 180 days (approximately July 13, 2026), the President may impose tariffs, quotas, import restrictions, or minimum import prices Federal Register; White House Fact Sheet.

Project Vault (February 4, 2026): A $12 billion public-private strategic stockpile, with a $10 billion EXIM Bank loan — reported as the largest in Export-Import Bank history — and $2 billion in private financing Chatham House; Conference Board.

FORGE Launch (February 4, 2026): The Forum on Resource Geostrategic Engagement, successor to the Minerals Security Partnership, launched at a Critical Minerals Ministerial attended by representatives from 54 countries and the European Commission, including 43 foreign and other ministers State Department. Chaired by South Korea until June 2026. Eleven new bilateral critical minerals frameworks or MOUs were signed with Argentina, the Cook Islands, Ecuador, Guinea, Morocco, Paraguay, Peru, the Philippines, the UAE, the UK, and Uzbekistan State Department.

The U.S. Government reported supporting projects with more than $30 billion in letters of interest, investments, loans, and other support over the preceding six months GlobalSecurity.org. Additional Defense Production Act Title III awards include $29.9 million to ElementUS Minerals for gallium and scandium separation, $23.4 million to 6K Additive for waste-to-metal upcycling, $90 million to Albemarle for the Kings Mountain lithium mine restart, and a $169 million DLA contract to Ares Strategic Mining for fluorspar. The DOE announced a $500 million funding opportunity for critical minerals processing, battery materials, and recycling capacity.

The Allied Producers section of most critical minerals analyses reads as a catalog of announced capacities and signed MOUs. This report takes a different approach, organizing the analysis around three strategic questions that determine whether the allied ecosystem can actually perform its assigned function: (a) Which allied producers can credibly address the heavy REE separation bottleneck — the decisive structural chokepoint — and on what timeline? (b) Which allied states face structural conflicts between U.S. alignment and Chinese economic ties that could undermine FORGE commitments at critical moments? (c) What is the minimum viable allied ecosystem for defense applications, and is it achievable by 2028?

On the heavy REE separation bottleneck: The most acute vulnerability in the U.S. supply chain is not the mining of light rare earth elements — Mountain Pass addresses that — but the separation and processing of heavy rare earth elements, particularly dysprosium and terbium, which are essential for high-temperature permanent magnets used in F-35 actuators, Virginia-class submarine systems, and advanced missile guidance. China holds a near-monopoly on this capability. Among allied producers, only a small number have meaningful near-term potential to address it. MP Materials is building heavy REE separation circuits at Mountain Pass using the $150 million DoD loan, but Mountain Pass ore is predominantly light REE bastnäsite with minimal heavy REE content — meaning the separation circuits will require third-party feedstocks whose provenance and China-exposure remain unresolved. REalloys in Canada is expanding heavy REE throughput by 300% (to approximately 30 MT dysprosium oxide and 15 MT terbium oxide annually) under cost-plus offtake agreements, with initial production targeted for 2027. This is meaningful for defense applications but modest relative to total demand. Lynas Rare Earths operates the most significant non-Chinese rare earth processing infrastructure outside China, with the Mt Weld mine and Kalgoorlie concentration in Australia and advanced separation at Gebeng, Malaysia, plus U.S. contracts for a light REE separation plant in Texas — but specific 2025–2026 heavy REE capacity updates are limited in available data, and Lynas's Malaysian operations have faced periodic political and regulatory pressure that introduces supply reliability risk. Japan's deep-sea mining project at Minamitorishima holds centuries-potential supply of heavy REEs but remains pre-commercial and is unlikely to contribute meaningfully before 2030. The honest assessment is that the heavy REE separation bottleneck will not be resolved by 2028; the goal for that timeframe is to reduce, not eliminate, dependence on Chinese processing for defense-critical applications.

On structural conflicts between U.S. alignment and Chinese economic ties: South Korea's role as FORGE chair until June 2026 is strategically significant precisely because it illustrates the tension the entire coalition faces. South Korea has a formal security alliance with the United States and deep anxiety about Chinese mineral leverage, but it also has the largest trade relationship with China of any FORGE member, with POSCO and LG deeply integrated into Chinese supply chains for battery materials and rare earth alloys. Seoul's willingness to chair FORGE signals genuine commitment, but its ability to enforce coordinated price floors or restrict Chinese-processed inputs will be tested when the economic costs become concrete. The same tension applies, in varying degrees, to Japan (which reduced Chinese REE import dependence from 85% to 58% over a decade but retains significant commercial exposure), Australia (which has navigated a bruising trade dispute with China while maintaining mineral export relationships), and virtually every resource-rich developing country in the coalition. The DRC, Guinea, Kazakhstan, and Peru all receive significant Chinese infrastructure investment and off-take agreements; their participation in FORGE frameworks does not preclude parallel engagement with Chinese buyers, and the U.S. should not assume exclusivity from bilateral MOUs that do not explicitly require it.

On the minimum viable allied ecosystem for defense applications: The defense sector's magnet demand is projected at approximately 5,000 tonnes by 2028. Against this benchmark, the emerging allied ecosystem — MP Materials' combined Independence and 10X facilities targeting approximately 10,000 MT/year by 2028, USA Rare Earth's Stillwater facility targeting 3,000 MT/year, Noveon's 1,000 MT/year recycling operation, and REalloys' expanded Canadian processing — could in principle cover defense requirements if it comes online on schedule and if heavy REE feedstock gaps are addressed. This is a meaningful but fragile sufficiency: it depends on multiple complex industrial projects executing on aggressive timelines simultaneously, and it leaves commercial demand — projected at 145,000 MT for e-mobility magnets alone by 2030 — almost entirely unaddressed by domestic and allied capacity. The minimum viable defense ecosystem is achievable by 2028 under optimistic assumptions; the minimum viable commercial ecosystem is not achievable within this decade.

MP Materials (Mountain Pass, California): The sole large-scale rare earth mine in the Western Hemisphere, producing approximately 40,000 metric tons per year of total rare earth oxides (TREO), with expansion plans toward 60,000 MT Payne Institute. The company has re-established light REE separation on-site and is building heavy REE separation circuits for dysprosium, terbium, and samarium with the $150 million DoD loan. Its Texas operations include the 1,000 MT/year Independence facility in Fort Worth (with GM contracted for output), expanding to 3,000 MT, and the 7,000 MT/year 10X facility at a 120-acre site in Northlake, Texas (groundbreaking February 27, 2026), targeting commissioning in 2028 Resource Recycling. Aggregate U.S. magnet capacity target: approximately 10,000 MT/year. The critical unresolved question is feedstock: Mountain Pass bastnäsite yields less than 20 MT of dysprosium from ore content, meaning heavy REE separation circuits will require third-party inputs whose supply security is not yet established.

USA Rare Earth (Stillwater, Oklahoma): Completed a 310,000 ft² sintered magnet Innovation Lab designed for 3,000 MT/year NdFeB output, with first commercial production slated for the first half of 2026 (single-source: industry analyses, not independently corroborated). The company's Round Top, Texas mine-to-magnet project is backed by a reported $1.6 billion U.S. government package ($277 million in incentives plus $1.3 billion loan) and $1.5 billion in private investment (single-source: industry analyses, not independently corroborated).

Noveon Magnetics (San Marcos, Texas): Scaling a 1,000 MT/year facility that upcycles end-of-life magnets into new NdFeB products, claiming 90% energy savings over conventional processing (single-source: company reporting, not independently corroborated). Noveon represents the most advanced U.S. recycling operation but at a scale that is a fraction of projected demand.

Lynas Rare Earths (Australia/Malaysia): Operates the Mt Weld mine and Kalgoorlie concentration in Australia, with advanced separation and refining at Gebeng, Malaysia. Has U.S. contracts for a light REE separation plant in Texas, though specific 2025–2026 capacity updates are limited in available data. Lynas is the most operationally mature non-Chinese rare earth processor, but its Malaysian operations introduce political risk, and its focus on light REEs means it does not directly address the heavy REE separation bottleneck.

REalloys (Canada): A $21 million expansion boosting heavy REE throughput 300% (to 30 MT dysprosium oxide, 15 MT terbium oxide) and NdPr metal 50% (to 400–600 MT/year), with initial production in 2027 under cost-plus offtake agreements. REalloys is the most significant near-term allied contribution to the heavy REE separation gap, though its scale remains modest relative to defense requirements.

Japan: Japan's post-2010 diversification — reducing Chinese REE import dependence from 85% to 58% over a decade — is the most instructive historical precedent for the current U.S. challenge. What made that diversification possible was a combination of government-backed offtake agreements with Lynas and other non-Chinese producers, aggressive investment in recycling and substitution technology, and a willingness to absorb above-market costs for supply security. The structural differences that make the current challenge harder include China's subsequent consolidation of midstream processing dominance (which was less complete in 2010), the broader commodity scope of current controls, and the shorter political timelines that democratic governments face. Japan's rare earth market is projected at USD 830.1 million in 2025, growing to USD 2,523.6 million by 2034 at 13.15% CAGR. Key players include Proterial (formerly Hitachi Metals) and Aichi Steel (Toyota subsidiary) in magnet production. Japan's deep-sea mining project at Minamitorishima holds centuries-potential supply but remains pre-commercial. A U.S.-Japan framework (October 2025) targets joint mining and processing for heavy and light REEs.

South Korea: Chairs FORGE until June 2026 — a role that signals genuine strategic commitment but also places Seoul in the uncomfortable position of managing an initiative that directly challenges its largest trading partner. POSCO and LG are active in REE magnets and alloys, with POSCO exploring upstream processing and recycling, though specific 2025 capacity figures are not publicly available. South Korea's ability to sustain FORGE leadership through the November 2026 truce expiration, when the coalition's commitments will face their first real test, is a leading indicator of the initiative's durability.

Corporate demand signals: Apple committed $500 million to source magnets manufactured in Texas (single-source: ORF America, not independently corroborated) ORF America, anchoring private-sector demand alongside DoD procurement. This commitment is strategically significant because it demonstrates that commercial demand pull — not just government procurement — can anchor the economics of domestic magnet manufacturing.

The 2025 crisis unfolded in three distinct phases, each calibrated to maximize leverage while preserving reversibility. The thread connecting all three phases is the licensed scarcity mechanism: at no point did China impose a formal embargo, and at every point it retained administrative discretion to throttle supply to specific buyers, end-uses, or countries without triggering the political mobilization that a formal cutoff would produce.

Phase 1 — The April Salvo: MOFCOM Announcement No. 18 (April 4, 2025) imposed export licensing on seven medium and heavy rare earth elements: samarium, gadolinium, terbium, dysprosium, lutetium, scandium, and yttrium, along with their oxides, alloys, compounds, mixtures, and certain permanent magnet materials Holland & Knight. The controls were announced the same day China imposed a symmetric 34% tax on U.S. imports, in direct response to President Trump's April 2, 2025 "Liberation Day" tariffs AP. The licensing regime applied globally, not just to U.S. buyers, complicating rerouting through third countries. Exporters must ensure end customers are not involved in sensitive sectors such as military hardware, giving Beijing maximum discretionary leverage to throttle supply to specific end-users without triggering a formal embargo — the licensed scarcity mechanism in its purest form.

Phase 2 — The October Expansion: MOFCOM Announcement No. 61 (October 2025) added five rare earths — erbium, europium, holmium, thulium, and ytterbium — and broadened scope to manufacturing equipment and technologies for rare earth separation and magnet production, plus some graphite and battery components CSIS. The inclusion of holmium was strategically calculated: many permanent magnet makers had been revising formulations to substitute holmium for previously restricted elements since April, and Beijing closed the escape route IEA. This adaptive closure of substitution pathways — responding to Western workarounds in near-real time — demonstrates the sophistication of China's control architecture and the difficulty of engineering around it.

Most consequentially, China deployed its own version of the U.S. Foreign Direct Product Rule (FDPR) — the first time Beijing applied this mechanism. Under the new rules, controls extraterritorially regulate Chinese-origin rare earths exported by foreign entities, foreign-made permanent magnet materials where Chinese content value reaches 0.1% or more (de minimis rule), and foreign-made items manufactured using Chinese rare earth mining, separation, smelting, magnet manufacturing, or recycling technology CM Trade Law. Starting December 1, 2025, companies with any affiliation to foreign militaries would be largely denied export licenses, and requests to use rare earths for military purposes would be automatically rejected CSIS.

This extraterritorial reach means that even if a U.S. company sources REEs from Australia, if those REEs were processed using Chinese technology, they may still fall under Beijing's licensing authority — a chokepoint that cannot be resolved by mining diversification alone. The technology-based extraterritorial control is arguably more consequential than the commodity-based controls, because it reaches into the processing infrastructure of allied nations and constrains the FORGE ecosystem's ability to build on Chinese-derived technical knowledge.

Phase 3 — The Truce: At the October 2025 Xi-Trump summit, Beijing agreed to a one-year suspension of the proposed October expansions, deferring most measures until approximately late 2026, while the U.S. rolled back tariffs Barnes & Thornburg. The precise expiration date of the October 2025 suspension is reported variously as October or November 2026 across available sources; this ambiguity should be resolved against the original MOFCOM suspension text before operational planning, as the difference has material implications for the Section 232 negotiation timeline and FORGE milestone sequencing. Critically, the April 2025 controls on the original seven elements remain fully in force regardless of the truce. The suspension did not remove underlying licensing requirements; exporters still need MOFCOM approval Pillsbury. The "pause" is best understood as a tactical de-escalation to gain diplomatic credit and blunt calls for decoupling, not a structural retreat.

The truce's fragility was demonstrated during the May 11, 2025 talks in Switzerland, where U.S. and Chinese officials agreed to a 90-day tariff truce that included removing U.S. companies from China's trade blacklist and restoring rare earth access. Yet U.S. manufacturers soon began shutting down production as China delayed export license issuance despite not formally abandoning the deal CSIS. This is the licensed scarcity dynamic in operational form: selectively slowing or denying approvals to particular firms, uses, or countries while formally remaining within WTO-compatible rules. It is the most dangerous form of Chinese leverage precisely because it is difficult to attribute, hard to challenge legally, and calibrated to stay below the threshold that would trigger a decisive Western response.

China's weaponization extends well beyond rare earths. The December 2024 opening salvo imposed export license requirements on gallium and germanium (indispensable for advanced semiconductors) and antimony (defense applications). New rules in 2026 restrict tungsten and antimony exports. China has elevated silver to strategic-material status, placing its export controls on the same regulatory footing as rare earths CNBC. U.S. federal critical mineral policy in 2026 is shifting focus beyond rare earths toward antimony and tungsten, given near-total dependence on China, Tajikistan, and Russia for defense alloys, munitions, and flame-retardant applications Alcircle. Over 200 mineral export restriction measures were imposed by various countries globally in 2025 Global Trade Alert. The expanding commodity front means that the U.S. strategy risks being perpetually reactive — addressing each mineral crisis as it emerges rather than building systemic resilience across the full critical minerals stack. This is a structural weakness in the current approach that no single bilateral MOU or stockpile commitment can resolve.

The most critical analytical finding is that the chokepoint is not mining but midstream processing. Increased mining is arguably the easy part; extraction will not address supply chain concerns if the ore still must be shipped to China to be processed and turned into batteries, magnets, and other products Global Policy Watch. China controls more than 90% of the downstream value chain, including oxide separation, metal refining, and magnet production. It holds a monopoly on the separation of dysprosium and terbium — critical for manufacturing permanent magnets capable of withstanding high temperatures in F-35 jet actuators, Tesla's humanoid robots, and Virginia-class submarines GQG Partners.

Mountain Pass ore is predominantly light REE bastnäsite, meaning heavy REE separation will require third-party feedstocks or recycled inputs — a dependency not yet fully resolved. MP Materials is only beginning to expand into heavy rare earth separation and magnet production at commercial scale Columbia SIPA/CGEP. Even at the planned 10,000 MT total U.S. magnet capacity by 2028, this would cover projected defense magnet demand (approximately 5,000 tonnes by 2028) but represent a fraction of total commercial requirements — U.S. e-mobility magnet demand alone is projected to reach 145,000 tonnes by 2030 Mining.com.

RFF's analysis raises a pointed challenge to the onshoring model: the $110/kg NdPr price floor is approximately $50/kg above spot price, creating a value chain that may not be internationally competitive without sustained government support RFF. China's ability to flood markets with below-cost production — as it did with solar panels and lithium — represents a long-term threat to commercial viability even after Western capacity is built. This is the core tension in the U.S. strategy: the security imperative demands domestic capacity, but the economic model requires either permanent subsidy or structural market reform (e.g., FORGE price floors enforced through coordinated tariffs). The Molycorp bankruptcy of 2015 is the cautionary precedent: a U.S. rare earth producer that reached commercial scale only to be driven into insolvency when China flooded the market with below-cost production. The DoD's 10-year price floor protects MP Materials against this scenario, but it does not protect the broader ecosystem of allied and domestic producers who lack equivalent guarantees.

The defense sector faces the most acute long-term risk. Even before the latest measures, the U.S. defense industrial base had limited production capacity and limited ability to rapidly scale CSIS. An academic study modeled a "strategic rare earth supply cut-off" as a non-kinetic deterrence tool against the U.S. military-industrial complex, emphasizing the outsized impact on specific weapons systems even with partial disruptions arXiv. The restrictions impact defense contractors, Apple, Nvidia, Intel, Tesla, Ford, and GM CNBC.

The National Defense Stockpile currently includes copper, nickel, lithium, antimony, and 16 rare earth elements. The 2025 procurement requests specifically targeted samarium, dysprosium, terbium, scandium, tungsten, and graphite. DoD announced intent to procure up to $1 billion in stockpile materials in 2025. The licensed scarcity mechanism is particularly dangerous for defense applications because it can be calibrated to target specific end-uses — denying licenses for military-adjacent applications while continuing to supply civilian buyers — in ways that are difficult to counter without a fully sovereign domestic supply chain.

The Council on Foreign Relations has articulated the most intellectually significant challenge to the mining-centric approach: the United States cannot out-mine and out-process China and should instead leapfrog China's dominance by scaling disruptive innovation, recovery, and recycling, which is cheaper, cleaner, and faster to deploy CFR. CFR recommends accelerating substitute materials — such as rare-earth-free and reduced-content magnets — to eliminate exposure to the most geopolitically vulnerable inputs. BMW is rolling out cars in 2026 that are rare earth free, having re-engineered their motors and magnets CFR event.

However, the global rare earth recycling market remains nascent: recycling production was less than 5 kilotonnes of REE in 2025, against targets to exceed 20 kilotonnes by 2035 and projected magnetic REE demand of 176 kilotonnes by 2035. Manufacturing scrap currently dominates recycler feedstocks, with end-of-life magnets projected to represent over 50% of recycling revenue only by 2032. Hydrometallurgical leaching processes achieve greater than 95% oxide recovery, and high-performance magnet separation (HPMS) technologies in Japan and Korea pursue closed-loop systems, but commercial scale remains years away. Noveon Magnetics' 1,000 MT/year San Marcos facility represents the most advanced U.S. recycling operation, but this is a fraction of demand.

The CFR view is analytically compelling but commercially unproven at scale. Private capital alone cannot reliably bridge the gaps facing frontier mineral technologies, particularly when timelines are long and returns depend on system-wide adoption CFR. Government support is a necessary catalyst, not a substitute for markets. The most realistic near-term contribution of the innovation pathway is not to replace the mining and processing buildout but to reduce the REE intensity of key applications — particularly in commercial automotive and consumer electronics — thereby shrinking the demand gap that domestic and allied supply chains must fill.

The five options analyzed below are not mutually exclusive; the report's preferred strategic posture, articulated at the end of this section, argues for a sequenced combination. They are presented separately to enable clear assessment of each option's strengths, weaknesses, and second-order effects.

Mechanism: Government equity stakes, price floors, long-term offtake agreements, and DPA Title III awards to build mine-to-magnet capacity on U.S. soil.

Current trajectory: MP Materials' phased buildout (Independence at 3,000 MT by ~2026, 10X at 7,000 MT by 2028) is the anchor. USA Rare Earth's Stillwater facility (3,000 MT) and Noveon's recycling plant (1,000 MT) add capacity. Mountain Pass expansion toward 60,000 MT TREO and new heavy REE separation circuits address upstream gaps.

Strengths: Provides sovereign supply for defense applications. DoD's 10-year, 7,000 MT/year purchase commitment and $110/kg price floor de-risk private investment and explicitly protect against the Molycorp-pattern price-dumping scenario. Apple's $500 million Texas magnet commitment (single-source: ORF America) demonstrates commercial demand pull. The Colosseum Rare Earth Project received DOI approval on April 8, 2025, potentially becoming the second U.S. rare earth mine adjacent to Mountain Pass.

Weaknesses: Timelines are the critical constraint. The 10X facility does not commission until 2028. Even at full capacity, 10,000 MT/year covers defense but is a fraction of projected commercial demand (145,000 MT for e-mobility alone by 2030). Heavy REE separation from Mountain Pass bastnäsite ore yields minimal dysprosium (less than 20 MT from ore content), requiring third-party feedstocks that may themselves be China-exposed. The $110/kg price floor is approximately double spot price, creating a permanently subsidized value chain — a point RFF highlights as potentially unsustainable RFF. Permitting timelines for new mines average 4–10 years for environmental impact statements under NEPA, though Mountain Pass leverages existing 30-year permits.

Second-order effects: Success would demonstrate that government market-making can unlock private capital for strategic industries, potentially creating a template for other critical minerals. Failure — through delays, cost overruns, or Chinese price dumping — would discredit the model and strengthen arguments for managed interdependence. The political economy of above-market mineral prices has not been tested: higher EV costs and defense procurement expenses will eventually generate domestic political resistance that could erode the subsidy commitments underpinning the model.

Mechanism: Plurilateral coalition with coordinated price floors, preferential trade zones, bilateral MOUs, and joint investment in mining and processing across allied and partner nations.

Current trajectory: FORGE succeeds the MSP with all 17 original members plus broader participation. Eleven new bilateral frameworks signed at the February 2026 ministerial. Specialized agreements leverage partner strengths: Australia for mining expertise in rare earths, lithium, and gallium; Saudi Arabia for heavy rare earth deposits and cheap energy for large-scale processing; the DRC for cobalt and copper (with a Glencore-Orion Critical Mineral Consortium MOU witnessed by Deputy Secretary Landau) State Department; GlobalSecurity.org. Pax Silica, announced December 2025, focuses on securing the semiconductor-adjacent mineral stack with nine partners and over 50 nations attending related events Wikipedia — Pax Silica.

Strengths: Diversifies supply across multiple geographies and political systems. Japan's post-2010 diversification reduced China's share of its REE imports from 85% to 58%, demonstrating feasibility over a decade. Vice President Vance's description of "reference prices for critical minerals at each stage of production" maintained through "adjustable tariffs to uphold pricing integrity" Atlantic Council signals a serious mechanism for countering Chinese price manipulation. If FORGE can operationalize coordinated price floors, it would address the single greatest commercial vulnerability of the domestic onshoring model.

Weaknesses: Framework agreements are not operational mines, as the Atlantic Council notes — each bilateral deal requires different concessions, obligations, and political risks Atlantic Council. Many resource-rich countries have deep existing relationships with China: the DRC, Guinea, Kazakhstan, and Peru all receive significant Chinese infrastructure investment and off-take agreements. Chinese commentators argue many FORGE participants would be reluctant to give up access to China's market Asia Times. Chinese state media has labeled the initiative "Metals NATO," revealing Beijing's concern but also its counter-narrative of "open, inclusive international trade" versus "exclusive blocs" Asia Times.

Chatham House raises the fundamental "commitment problem": the U.S. strategy explicitly aims not to overtake China in production volumes but to create a new geopolitically exclusionary mining and processing ecosystem. Success requires creating bureaucracies, departments, and agencies that generate a web of vested interests around implementation — interests that become very hard for successor administrations to undo Chatham House. Whether the Trump administration's hub-and-spoke bilateral architecture can achieve this institutional durability is an open question. The 180-day Section 232 negotiation deadline (approximately July 13, 2026) is the first real test of whether FORGE members will accept binding price floor commitments or retreat to non-binding expressions of intent.

The Ukraine minerals dimension adds geopolitical complexity: a 50/50 partnership framework is under discussion for Ukraine's titanium, lithium, and rare earth deposits, but some Ukrainian commentators see the minerals deal as economic partnership devoid of hard security commitments Wikipedia — Ukraine–U.S. Minerals Agreement.

Mechanism: $12 billion public-private reserve to buffer short-term supply shocks, with EXIM Bank financing and private-sector participation.

Strengths: Provides immediate insurance against supply disruptions. Can be deployed within months rather than the years required for new processing capacity. Signals seriousness to both allies and adversaries. Stockpiling is the only option that addresses the vulnerability window between now and 2028, when domestic and allied processing capacity begins to come online. It is therefore a necessary complement to, not a substitute for, the longer-term structural options.

Weaknesses: Stockpiles buffer shocks of months to a couple of years but cannot substitute for industrial capacity. Public data on Project Vault's commodity breakdown, release triggers, and rotation rules remain sparse, limiting assessment of how long the U.S. could ride out a severe REE supply shock. The World Economic Forum and other analysts question long-term utility, noting that against China's 90% control of the magnet value chain, stockpiling may merely delay pain without solving structural dependency. The $12 billion figure, while large, must cover multiple minerals across defense and commercial applications — the per-mineral buffer may be thinner than headline numbers suggest. Critically, stockpiling does not address the licensed scarcity mechanism: if China throttles approvals gradually rather than imposing a sudden cutoff, the stockpile may be drawn down incrementally without triggering the release protocols that would justify its deployment.

Mechanism: R&D investment in rare-earth-free magnets, advanced recycling, and material substitution to reduce REE intensity of key applications.

Strengths: Addresses the root cause of vulnerability rather than replicating China's supply chain at higher cost. BMW's 2026 rare-earth-free vehicle motors demonstrate commercial viability for at least some applications. DOE's $500 million funding opportunity and $134 million for tailings/e-waste recovery support the pipeline. Noveon's 90% energy savings over conventional processing suggest recycling can be economically competitive. Unlike the mining and processing buildout, innovation investments do not require sustained government price support to remain viable once technologies reach commercial scale.

Weaknesses: Technology readiness levels vary widely. Rare-earth-free magnets currently sacrifice performance in high-temperature applications critical for defense (jet actuators, submarine systems). Recycling production was less than 5 kilotonnes in 2025 against projected demand of 176 kilotonnes for magnetic REEs by 2035. The CFR acknowledges that private capital alone cannot bridge the gaps, and government support is a necessary catalyst CFR. The innovation pathway is most credible as a medium-to-long-term complement to the onshoring and allied supply chain options, not as a near-term substitute.

Mechanism: Bilateral or multilateral agreements with China establishing supply guarantees, transparency mechanisms, and mutual restraint on weaponization.

Strengths: PIIE and some trade economists argue the escalating mineral trade war is fundamentally lose-lose, imposing high costs on both sides and encouraging inefficient duplication of capacity. Managed interdependence — with negotiations focused on mutual supply chain transparency and reliability — could avoid self-inflicted shortages while preserving the efficiency gains of specialization. The RFF analysis supports this view, arguing that China's export controls are designed for temporary leverage, not permanent withdrawal, and that a negotiated framework could stabilize supply while both sides build alternative capacity. The economic case is genuine: the $110/kg price floor, the FORGE coordination costs, and the above-market pricing that will flow through to EV consumers and defense procurement all represent real efficiency losses that a negotiated framework could reduce.

Weaknesses: The security-oriented counterargument is compelling and deserves full engagement rather than dismissal. CSIS and Jamestown analysts argue that the U.S. has systematically under-priced the strategic risk of dependence on an authoritarian rival controlling critical inputs, and that the current crisis — however costly — is the necessary catalyst for a structural correction that should have begun a decade ago CSIS; Jamestown. The May 2025 Swiss talks demonstrated that even formal agreements do not prevent licensed scarcity — China delayed license issuance despite not abandoning the deal. This is the most damaging evidence against Option 5: if China cannot be relied upon to honor the administrative implementation of a formal agreement, the value of any negotiated framework is fundamentally compromised. China's amendment of its Foreign Trade Law to expand countermeasure authorities outside WTO mechanisms further erodes the rules-based framework that would underpin such agreements. The honest assessment is that managed interdependence is a desirable endpoint for civilian applications where the security stakes are lower, but it cannot be the primary strategy for defense-critical minerals as long as China retains the licensed scarcity mechanism and the political will to use it.

Mountain Pass operates under a 30-year permit secured in 2010, and the Colosseum project received DOI approval in April 2025 after years of delay due to its location in Mojave National Preserve. But new U.S. mines face NEPA review timelines of 4–10 years for environmental impact statements. California's stringent environmental regulations add complexity. MP Materials reports a lower environmental footprint than Chinese mines per a 2025 Life Cycle Assessment, but domestic opposition to new mining on environmental and Indigenous rights grounds constrains how far "friend-shoring" can go. This is a structural constraint that executive orders cannot fully override without legislative reform.

Major Western OEMs in automotive, aerospace, and electronics may prefer cheap Chinese supply over resilience unless forced by regulation or security classification. Their lobbying could dilute strict implementation of Section 232 remedies or FORGE-aligned procurement rules. This represents an underappreciated internal constraint: the defense industrial base is a captive buyer, but the commercial sector — which represents the vast majority of magnet demand — will resist above-market pricing absent regulatory compulsion. The licensed scarcity mechanism is particularly effective at exploiting this divide: by maintaining supply to commercial buyers while throttling defense-adjacent applications, China can prevent the commercial sector from becoming a political constituency for supply chain resilience.

The analysis above supports a sequenced, four-track posture that combines elements of all five options while being explicit about their respective roles and timelines.

In the near term (2026–2027), the priority is bridging the vulnerability window through aggressive stockpiling under Project Vault and accelerating the domestic onshoring buildout, with DoD procurement commitments and price floors providing the demand anchor. The $110/kg price floor and 10-year purchase commitment for MP Materials should be treated as the template for extending similar guarantees to USA Rare Earth, REalloys, and other allied producers who currently lack equivalent protection against Chinese price dumping. The 180-day Section 232 negotiation deadline (approximately July 13, 2026) should be used to extract binding price floor commitments from FORGE members — not merely expressions of intent — with the credible threat of tariffs on non-aligned imports as the enforcement mechanism.

In the medium term (2027–2028), the priority shifts to operationalizing the allied supply chain through FORGE, with particular focus on the heavy REE separation bottleneck. The minimum viable defense ecosystem — approximately 14,000 MT/year of combined domestic and allied magnet capacity, with sovereign heavy REE separation for dysprosium and terbium — should be the explicit planning target for 2028. Innovation investments in rare-earth-free magnets and recycling should be scaled aggressively for commercial applications, where the security stakes are lower and the substitution economics are more favorable.

In the long term (2029+), narrow, enforceable guardrails with China around civilian applications — where managed interdependence is economically rational and the security risk is manageable — can be pursued without compromising the sovereign defense supply chain. This is not a concession to Chinese leverage; it is a recognition that the goal is a "China-light" ecosystem, not zero interdependence, and that preserving some commercial relationship with China may be necessary to prevent Beijing from treating the entire FORGE initiative as an existential threat requiring maximum escalation.

This posture should be revised if: (a) China escalates to a formal embargo rather than licensed scarcity, which would require emergency mobilization of all available options simultaneously; (b) the 10X facility commissioning slips beyond 2029, which would extend the vulnerability window and require additional stockpile investment; or (c) FORGE members defect from price floor commitments under Chinese economic pressure, which would require a fundamental reassessment of the allied supply chain strategy.

The one-year suspension of the October 2025 controls expires in approximately late 2026 — with the precise date reported variously as October or November 2026 across available sources, an ambiguity that should be resolved against the original MOFCOM suspension text before operational planning. Whether the Xi-Trump rapport holds, and whether sufficient progress on FORGE and domestic capacity will be demonstrated to justify extension, is the single most consequential near-term uncertainty. The April 2025 controls remain regardless. TD Economics assesses that replacing Chinese supplies will likely be nearly impossible before the one-year truce expires TD Economics. Indicators to watch: MOFCOM licensing patterns, internal industry guidance, political context (Taiwan tensions, U.S. midterm dynamics). The licensed scarcity mechanism means that escalation need not take the form of a formal announcement — a gradual increase in license denial rates or processing delays would achieve the same effect with less political visibility.

China can flood markets with below-cost production to undermine the commercial viability of Western producers, as it did after the 2010 Senkaku dispute and as it did with solar panels and lithium batteries. The $110/kg NdPr price floor protects MP Materials for 10 years, but other producers (USA Rare Earth, Noveon, allied facilities) lack equivalent guarantees. If China drops NdPr prices to $40–50/kg, unprotected Western producers could face bankruptcy before reaching commercial scale — repeating the Molycorp pattern of 2015. This risk is not hypothetical: China has demonstrated both the capability and the willingness to deploy below-cost production as a strategic weapon, and the current buildout of Western capacity gives it strong incentives to do so before that capacity becomes self-sustaining.

FORGE's 54-nation ministerial attendance does not equal binding commitment. The 180-day Section 232 negotiation deadline (approximately July 2026) is the first real test. Key questions: Will the EU, Japan, and South Korea align on coordinated tariffs against Chinese magnets, or hedge to preserve Chinese market access? Will resource states like the DRC and Guinea resist exclusive alignment with U.S.-led schemes given deep Chinese economic ties? Argentina's balancing act between Western initiatives and Chinese investment illustrates the tension. The licensed scarcity mechanism is relevant here too: China can selectively maintain supply to wavering FORGE members while throttling supply to committed ones, creating economic incentives for defection that the coalition's institutional architecture must be designed to resist.

Announced capacities are ambitious; actual timelines for complex industrial projects routinely slip. The 10X facility's 2028 commissioning date is aggressive for a greenfield magnet manufacturing campus. Heavy REE separation at Mountain Pass using third-party feedstocks is technically unproven at commercial scale. If delays push meaningful capacity to 2029–2030, the vulnerability window extends dangerously, and the stockpile buffer — already thin on a per-mineral basis — becomes the primary line of defense.

China's extension of controls to silver, tungsten, antimony, graphite, and battery components signals that the weaponization playbook is not limited to REEs. Rising AI data center electricity demand creates pressure on aluminum, copper, titanium, and magnesium Alcircle. The U.S. strategy risks being perpetually reactive — addressing each mineral crisis as it emerges rather than building systemic resilience. The licensed scarcity mechanism can be applied to any commodity subject to export licensing, meaning the administrative infrastructure China has built for rare earths is readily extensible to the full critical minerals stack.

A cross-strait crisis would simultaneously trigger rare earth supply disruption (China controls the supply) and semiconductor supply disruption (TSMC controls 63% of advanced process capacity at 7nm and below, with 80–90% of AI server chips Taiwan-centric). TSMC's Arizona fab (3nm by 2026) and Japan/Kumamoto plant (2025) diversify 10–20% of capacity, but a blockade could halt 90% of advanced nodes per U.S. assessments. The rare earth and semiconductor vulnerabilities are mutually reinforcing in a Taiwan scenario: the same crisis that triggers REE supply disruption would also disrupt the semiconductor supply chain that depends on REE-derived materials, creating compounding effects that no single-domain analysis captures. This cross-domain linkage is underrepresented in most policy discussions and should be a central consideration in contingency planning.

The economic pain of above-market mineral prices, higher EV costs (potentially 10–20% short-term without diversification, inferred from supply chain trends), and defense procurement cost increases has not been politically tested. A prolonged mineral trade war could inflict severe damage on all parties, and the domestic political sustainability of these costs in the U.S. remains uncertain. The licensed scarcity mechanism is designed in part to exploit this vulnerability: by maintaining enough supply to keep consumer prices elevated but not catastrophic, China can sustain economic pressure over a timeframe that tests democratic governments' political will without triggering the emergency mobilization that a sudden cutoff would produce.

Reports of Chinese firm resistance to in-country processing mandates and local government pushback against tight export controls suggest potential fissures in Beijing's strategy. If enforcement is lax, this could create windows for Western buyers to access Chinese-processed materials through informal channels. However, this is an obscure variable with no reliable open-source verification, and should not be weighted heavily in planning. It is noted here for completeness and as a potential avenue for further intelligence collection, not as a basis for strategic assumptions.

The absence of formal WTO challenges to China's 2024–2025 controls reflects Appellate Body dysfunction and strategic calculation, not legal weakness — the 2014 DS431 precedent established clear violations. Future political shifts could reopen this avenue. The evolving interpretation of Article XXI national security exceptions will shape how far states feel free to weaponize minerals trade. However, given the multi-year timelines for WTO adjudication and the strategic calculation that filing a challenge signals weakness and constrains the complainant's own retaliatory options, formal WTO challenges are unlikely to be a primary instrument of U.S. strategy in the near term.

The truce buys time but does not resolve the structural problem. Two hard decision points converge: the 180-day Section 232 negotiation deadline (approximately July 13, 2026) and the late 2026 truce expiration (precise date to be confirmed against original MOFCOM suspension text). The U.S. will use both as leverage to accelerate FORGE commitments and bilateral deals. China will use the same timeline to extract concessions on tariffs and technology access. The greatest near-term risk is not a total cutoff but continued licensed scarcity — selective throttling that degrades supply without triggering the political mobilization of a formal embargo. This mechanism is designed to operate below the threshold of decisive response, and countering it requires not just supply chain diversification but the administrative and intelligence capacity to detect and attribute selective license denials in near-real time.

Confidence: High that the April 2025 controls remain in force and create ongoing supply friction. Medium that the truce holds through late 2026 — contingent on broader U.S.-China relations and absence of a Taiwan crisis.

The MP Materials 10X facility commissioning in 2028 is the most consequential single event in the U.S. domestic buildout. If it comes online on schedule and at projected capacity, it will represent the first genuinely sovereign U.S. mine-to-magnet supply chain in decades. Combined with USA Rare Earth's Stillwater facility and Noveon's recycling operations, total U.S. magnet capacity could reach approximately 14,000 MT/year — sufficient for defense and a meaningful fraction of commercial demand. Allied capacity additions (Lynas, REalloys, Japanese and South Korean producers) would further diversify the ecosystem.

However, heavy REE separation remains a bottleneck. Mountain Pass bastnäsite yields minimal dysprosium and terbium; third-party feedstocks and recycling must fill the gap. China's 94% share of sintered permanent magnet production will not be materially challenged by 2028 — the goal is resilience, not parity. The licensed scarcity mechanism will remain available to China throughout this period, and the 2027–2028 window — when Western capacity is under construction but not yet operational — is the period of maximum vulnerability to calibrated supply disruption.

Confidence: Medium that MP Materials achieves 10X commissioning on the 2028 timeline. Low that heavy REE separation achieves commercial scale without continued Chinese feedstock exposure.

Complete independence is unlikely, especially for heavy rare earth elements. The realistic endpoint is managed, diversified interdependence — shifting from unilateral vulnerability to shared, manageable risk. If FORGE delivers on coordinated price floors and processing investments, and if innovation in rare-earth-free magnets and recycling scales as CFR advocates, material Chinese leverage should diminish for defense applications and some commercial uses.

Ironically, China's weaponization of concentration may become the trigger that unwinds its dominance. By demonstrating the costs of dependence, Beijing accelerated diversification efforts that many democracies had treated as optional ORF America. The most durable path combines four parallel tracks: (1) domestic onshoring anchored by the DoD-MP Materials model, with price floor guarantees extended to the broader domestic producer ecosystem; (2) allied supply chain development through FORGE with binding price floor commitments enforced through coordinated tariffs; (3) innovation-led substitution and recycling, prioritized for commercial applications where rare-earth-free alternatives are technically feasible; and (4) narrow, enforceable guardrails with China around civilian uses where managed interdependence is economically rational and the security risk is manageable.

Whether the political will, financial commitment, and alliance cohesion can be sustained through the inevitable periods