The Critical Frontier: Technology’s New Battle for Minerals

Empires don’t rise on ideology alone - they rise by controlling the materials and trade routes that power their economies. The British Empire and its East India Company, for example, did not merely trade goods but systematically controlled the sourcing, processing, and distribution of spices, textiles, tea, and metals. By owning the arteries of global commerce, Britain translated material access into industrial advantage, military strength, and financial power. This is not isolated to the British Empire; for millennia, empires have endured by securing and maintaining the supply chains behind critical resources. 

The era of liberal globalization, however, has eroded this focus with its emphasis on specialization and efficient markets, extended value chains, and decades of relative geopolitical stability. The result is an assumption that access, not ownership, was sufficient. 

That assumption is now dissipating. Geopolitical fragmentation, the ongoing trade war, and mounting concerns around national security have pushed resource security to the forefront of strategic planning. 

A key battleground today centers largely on the minerals and processing capabilities that enable advanced industry: lithium for batteries, nickel and graphite for energy storage, rare earths for motors and defense technologies, and specialty materials for semiconductors. Not to mention more “household” metals such as copper and aluminum that are required in almost every product we use, but whose processing prowess sits largely outside of America. Access to these materials has quietly become the new upstream chokepoint for US industrial competitiveness. As the country pushes forward with reindustrialization, we’re seeing a shift from globalized sourcing as default to a renewed emphasis on sovereign control over strategic materials. 

What makes this moment different from prior eras of resource competition is the role of technology. Control over natural resources is no longer limited to what can be extracted from the ground; it increasingly includes what can be simulated, optimized, or even created in a lab. From AI-driven discovery to synthetic or lab-created materials, the frontier of resource security is beginning to blur the line between geology and technology - at times reading more like science fiction than traditional mining. This shift doesn’t replace natural resources, but it fundamentally reshapes how scarcity, control, and advantage are defined in the modern industrial economy.

The Critical Mineral Value Chain

The value chain can be broken down into the three segments: upstream, midstream, and downstream. Examples of companies that dominate each segment include the following. It’s worth noting that many of the “majors” such as BHP and Rio Tinto, touch multiple pieces of this value chain, as vertical integration unlocks favorable economics. 

Critical Minerals Value Chain Market Map

Note: Gray shaded boxes include examples of startups serving / disrupting that subsegment

1. Upstream: Exploration & Extraction

The upstream stage is all about finding and pulling minerals out of the ground, but the nuance lies in where and how. While many critical minerals are abundant in the Earth’s crust (e.g., copper, graphite, and more), economically viable deposits are far more rare. Ore grade, geology, permitting timelines, water access, community sentiment, and infrastructure all determine whether a deposit can actually become an economically viable mine. 

This stage carries significant capital requirements and long lead times, which is why upstream development tends to be dominated by countries and companies willing to take on regulatory friction and commodity risk. For deposits within the US, upstream is less about geology and more about the ability to permit and build assets fast enough to meet rising demand.

2. Midstream: Processing, Refining & Separation

Midstream is where raw material becomes usable material - and it’s the most severe bottleneck in the critical minerals ecosystem. Processing requires complex chemistry, specialized equipment, and decades of tacit know-how. Turning ore into battery-grade lithium carbonate, separated rare earth oxides, or high-purity graphite is far more challenging than mining itself. 

Rare earth elements - a subset of critical minerals - are actually not geologically rare. Rather, they are rare in a concentration that makes it viable to extract and also chemically complex to process into usable materials. This difficulty is precisely what makes them so strategically sensitive. Rare earths underpin the high-performance magnets and components that power electrification, defense systems, automation, semiconductors, and the fast-growing category of physical AI (aka robotics).

Aluminum illustrates how midstream capacity - not raw resource availability - can become a binding constraint. While bauxite is abundant globally, converting it into alumina and then primary aluminum requires enormous, continuous energy input, making smelting capacity highly sensitive to electricity prices and grid reliability. 

The midstream is also where geopolitical concentration is most severe: China controls the majority of global refining capacity, giving it leverage over downstream industries from semiconductors to defense systems. Midstream is capital-intensive and operationally demanding, but it’s also where the most strategic value accrues. Without domestic processing capacity, upstream exploration & extraction offer limited national or industrial benefit.

Share of Refined Material Production by Country

Source: IEA

3. Downstream: Component Manufacturing & Industrial Inputs

Downstream is where processed materials become components that make modern industry possible: magnets for electric motors and drones, cathodes and anodes for batteries, wafers for semiconductors, and alloyed metals for everything from turbines to robots. This stage of production is closest to end markets and therefore carries the strongest link to industrial competitiveness. A country may source minerals globally, but if it cannot turn them into finished components, it remains dependent on foreign manufacturing ecosystems. In sectors like magnets and battery materials, the US currently has minimal domestic production despite growing demand - creating an urgency to build downstream capacity alongside upstream and midstream investments.

The Impact of Scale

Scale is the defining feature of the critical minerals industry and the primary reason the US struggles to compete across extraction, processing, and component manufacturing. Unlike software - or even traditional manufacturing - this value chain rewards players who can deploy massive, long-duration capital and operate facilities at volumes that drive down unit costs and accelerate learning curves. As a friend at BHP Billiton reminded us, exploration campaigns routinely require tens of millions before a single economic deposit is confirmed; midstream refineries demand billion-dollar commitments before producing battery- or magnet-grade material; and downstream manufacturers only become cost-competitive when operating at volumes measured in tens of thousands of tons per year. 

These dynamics explain why countries with the willingness and balance sheet to invest at scale - most notably China - have built near-unassailable advantages. China recognized this opportunity decades ago and used its Belt and Road Initiative (“BRI”) to secure upstream resources, lock in offtake agreements, and finance infrastructure in mineral-rich countries - effectively underwriting the global scale needed to feed its midstream and downstream dominance. BRI didn’t just extend China’s geopolitical influence; it built the scaffolding for the industrial scale that now underpins its advantage in critical minerals.

The question of scale impacts each primary component of the value chain in the following ways:

Upstream

Exploration & extraction is a scale-driven business with long odds and long timelines. The average cost of discovering a new deposit has risen >300% on average over the past two decades, and less than 1% of mineral exploration projects result in actual mining production. Even reaching a feasibility study can require $5–50M in upfront exploration spending, and new mines often require billions to bring them fully online. These economics naturally favor players with large balance sheets or sovereign backing - those who can withstand decade-long development cycles and significant geologic risk.

In mining, scale creates a sharp stratification between juniors, mid-tiers, and majors - each playing a fundamentally different role in exploration. Junior miners are capital-light and high-risk, often operating with sub-$50M market caps and focused on early-stage exploration where technical insight and risk tolerance matter more than balance sheet strength; most never reach production and instead aim to sell assets upstream. Mid-tier miners bridge discovery and development, bringing projects through feasibility and permitting, but remain constrained by capital intensity and exposure to single-asset risk. Majors, by contrast, use scale as a competitive advantage: diversified portfolios allow them to absorb exploration failure, fund multi-year programs, deploy proprietary geological data, and ultimately decide which discoveries reach production. 

As a result, while innovation often starts with juniors, the economics of exploration - and the path to real supply - are ultimately dictated by the balance sheets and operating scale of the majors. This is something our friends at the majors frequently remind us of, but also complicates the path of a founding team going from startup to scale-up on venture timelines.

Midstream

If upstream requires scale, midstream demands it. Processing facilities typically require $1-3B in capex with a 7-12 timeline to return. Cost curves only become competitive at high throughputs - lithium conversion facilities need at least 5,000 tons/year to be viable, preferably 20,000+ tons/year for attractive economics. This is why China’s midstream dominance is so durable: it operates substantially larger than that of its Western peers and controls ~90% of global rare earth processing and ~99% of the three kinds of rare earths necessary for heat-resistant magnets.

In the midstream, scale determines whether processing economics work at all. Smaller “junior” players often emerge with novel chemistry or modular designs, but rarely reach cost competitiveness without operating at high, continuous throughput; failure here is less about technical feasibility and more about capital and execution. Mid-market operators may run one or two commercial facilities and anchor regional supply, but remain exposed to feedstock variability, energy costs, and customer concentration. True midstream majors differentiate through chemistry at scale - operating multiple high-throughput plants, absorbing commissioning risk, continuously improving yields, and leveraging integration across upstream feedstock and downstream offtake - making their cost advantage durable and difficult to displace.

Downstream

Downstream competitiveness is directly shaped by the scale of upstream and midstream supply. China’s JL Mag produces 29,300 tons/year of Neodymium Iron Boron (“NdFeB”) magnets, the world’s strongest, while MP Materials, the leading US rare earth materials company, only produces 1,000 tons/year. Without comparable scale feeding into the system, downstream manufacturers in the US face higher input costs, constrained capacity, and structural disadvantages in global markets.

The stratification within the downstream is driven by volume, reliability, and customer trust rather than geology or chemistry. Smaller manufacturers can innovate around performance or process improvements, but often struggle to meet the scale, consistency, and pricing demanded by large OEMs without significant capital investment. Mid-market players serve niche or regional demand, balancing specialization against rising cost pressure. Majors, by contrast, dominate through sheer throughput and learning curves - producing at volumes that lower unit costs, accelerate iteration, and lock them into OEM supply chains - creating defensibility that compounds over time and makes displacement exceedingly difficult.

Incumbent Software Systems

Across the critical minerals value chain, most operators rely on software systems that were designed for a slower, more stable industrial world. In upstream exploration and mining, tools excel at geological modeling and mine planning - but they leave users stitching together fragmented datasets, relying heavily on expert intuition, and waiting months for insights that increasingly need to be made in weeks. Geologists and operators spend more time reconciling drill logs, surveys, assays, and satellite data than testing new hypotheses. The result is slow iteration, limited probabilistic decision-making, and high dependence on scarce human expertise at precisely the moment discovery risk and capital intensity are rising.

Midstream processing exposes a different set of frustrations. Chemical modeling and plant automation systems perform well once processes are known and stable, but struggle during commissioning, ramp-up, and periods of feedstock variability - exactly when errors are most costly. Engineers lack real-time visibility into why yields fluctuate, why energy intensity spikes, or why scale-up timelines slip. Optimization happens locally, within individual units or plants, rather than across the full system, leaving operators to absorb commissioning risk manually while billion-dollar facilities sit below capacity.

Downstream manufacturers face yet another mismatch. ERP, MES, and factory automation systems track production, inventory, and quality after the fact, but offer little support for anticipating material variability or supply risk upstream. When input quality changes - whether in magnets, battery materials, or alloys - operators discover the problem only after yields fall or defects rise. These systems excel at reporting what happened, but leave users reactive in a world that increasingly demands foresight.

Across all segments, compliance and provenance tools add friction rather than clarity. Documentation remains paper- and PDF-heavy, verification is episodic rather than continuous, and users are forced to manage growing regulatory, labor, and trade requirements through manual processes and intermediaries. As critical minerals shift from commodities to strategic assets, this patchwork of legacy software becomes a structural constraint - slowing decisions, obscuring risk, and increasing operational exposure at every stage of the value chain.

These gaps create room for new generations of AI-native systems of intelligence that can reason across uncertainty, scale decisions in real time, and orchestrate complex industrial workflows end-to-end.

Summary of Technologies Used in the Sector

Potential investment opportunities

Next-generation exploration technologies can alleviate the long discovery timelines, which often stretch a decade, and address the discovery-to-mine success rate of 0.5% for greenfield sites and ~5% for brownfield sites. There is an opportunity for AI-native platforms that combine basin-scale simulation, historical drilling data, geophysics, and probabilistic modeling to identify drill targets. Winning solutions help decide where not to deploy capital as much as where to drill - compressing multi-year exploration funnels into fewer, higher-conviction bets. These tools also stabilize portfolio economics in an industry notorious for boom-and-bust cycles by improving hit rates and accelerating time-to-value.

Advanced processing and separation technologies can address the primary bottleneck of the value chain, the midstream, which is the hardest part of the supply chain precisely because it’s a chemistry + capex equation, not a software one. Rare earth separation plants, copper smelting facilities, and graphite purification lines often exceed hundreds of millions of dollars and take years to commission - creating enormous execution risk. Startups that can reduce capex, improve yield, optimize solvent usage, or accelerate commissioning timelines have the opportunity to reshape cost curves and reduce dependence on China. Whether through AI-driven sorting, membrane-based separation, or modular processing units, innovations here directly determine national competitiveness.

Circularity and advanced recycling platforms offer a way around permitting timelines and commodity cycles by tapping “above-ground mines” containing high concentrations of critical minerals. Automated disassembly, hydrometallurgical recovery, and AI-driven process optimization can make recycled inputs competitive with virgin materials - critical in markets where midstream scaling is slow. 

Provenance and compliance infrastructure can help address the tightening enforcement around labor, environmental standards, and trade restrictions for minerals, which will require the same level of traceability as food, pharmaceuticals, and industrial parts with impending compliance regimes. We see an opportunity for both software and hardware approaches, where digital traceability platforms are reinforced by physical markers. For example, we’ve had portfolio companies and third parties alike use certain mineral compositions or materials like diamond dust to demonstrate authenticity. Together, these systems enable downstream authentication, protect proprietary materials and processes, and make counterfeiting economically unviable in high-value supply chains.

Industrial-scale automation solutions for mineral processing facilities can help address their extremely complex operations with high failure rates, long commissioning periods, and limited visibility into process performance. Automation layers - ranging from sensor fusion to predictive maintenance to agentic control systems - can improve throughput, reduce reagent usage, and cut downtime. 

Tech-enabled manufacturing for rare earth components can address a production process that is energy-intensive, yield-sensitive, and highly dependent on tacit know-how. Technology layers ranging from advanced automation and in-line quality sensing to process control and software-driven yield optimization can improve consistency, reduce energy intensity, and shorten ramp-up timelines. Applied at scale, these systems create a path to domestic, lower-carbon magnet production for robotics, defense, and other high-performance applications.

Predictive maintenance for continuous, mission-critical assets can mitigate the expense associated with the downtime of continuously operating mining and processing facilities. Predictive maintenance platforms that integrate vibration, thermal, acoustic, and process data can prevent cascading failures across crushers, mills, conveyors, and refineries. The opportunity lies in moving from alerting to prescriptive, agentic workflows that recommend or trigger maintenance actions in real time.

Ultimately, US reindustrialization will be won by control over the physical systems that convert raw materials into strategic inputs at scale. Across critical minerals, the bottlenecks are more technical than geological, creating fertile ground for technology-driven disruption. We see compelling opportunities in companies that bring software, automation, and advanced manufacturing into historically analog, capital-heavy parts of the value chain. As critical minerals shift from commodities to strategic assets, the firms that modernize how these materials are discovered, processed, and protected will sit at the center of the next industrial build cycle - and represent the types of durable, systems-level investments we seek to back.