Each stage of raw material extraction leaves an irreversible mark on landscapes, human and more-than-human beings.

Locating the Minerals
AI begins with geologists, drones, and satellites scanning landscapes for mineral-rich sites. Exploration sounds harmless, but it often disturbs Indigenous lands, disrupts fragile ecosystems, and triggers conflicts over land rights.

Claiming the Land
Once a potential deposit is identified, mining companies stake claims and obtain necessary permits, ensuring legal rights to extract resources and compliance with environmental regulations. This often happens often in countries where environmental protections are weak and community voices are hardly listened to.

Infrastructure Takeover
Mining requires roads, processing plants, and worker camps, transforming untouched landscapes into industrial zones. These projects often displace communities and destroy local ecosystems before a single mineral is extracted.

Extraction: Open-Pit and Underground Mining
Depending on the deposit's location and type, minerals are extracted using methods like open-pit mining (vast, gaping holes in the earth) or underground mining (deep, dangerous tunnels). Both are energy-intensive, water-depleting, and highly polluting.

Transportation
After extraction, the raw materials are transported--often over great distances by truck, train, or ship--to refining facilities, a process that increases carbon emissions, raises costs, and can disrupt communities and ecosystems along transit routes.

Refining the Materials
Raw minerals must be purified before they can be used in AI hardware. Extracted ores undergo crushing, grinding, and separation processes to concentrate the valuable minerals. Techniques such as flotation, magnetic separation, and leaching are employed to achieve the desired purity. This step is further analyzed in our Research Area Materials Manufacturing.

The refining and processing of raw materials into components usable for AI systems introduces a different, though interconnected, layer of environmental and social complexity. While this stage is often perceived as more technologically advanced or contained than extraction, it involves equally significant transformations—both in terms of material impact and power relations.

High-temperature processes, chemical treatments, and large-scale water use define many aspects of materials manufacturing, but their consequences unfold unevenly depending on where production occurs, and under what regulatory conditions. This creates a spatial distribution of harm in which exposure to industrial waste, water scarcity, or air pollution may be concentrated in particular zones, often distant from where technological benefits are ultimately realized. Yet the effects are not always immediate or dramatic; they may accumulate gradually, becoming visible only over time through shifts in public health, land viability, or ecological resilience.

From a planetary justice perspective, the core concern is not only the environmental cost of these processes, but also how decisions about manufacturing are made—and who is able to participate in shaping them. The concentration of refining infrastructure in select regions often reflects global hierarchies of labor, capital, and influence, with limited avenues for those most affected to intervene in decisions around siting, emissions, or mitigation.

This stage also underscores how value is created in AI supply chains: as raw materials are transformed into high-tech inputs, control and profit tend to consolidate further upstream. A just approach to materials manufacturing would require broader engagement with questions of distributive and procedural fairness, ensuring that environmental responsibility is not treated as an externality and that the costs of technological advancement are not disproportionately borne by the same regions repeatedly exposed to extractive harm.

Mining for these materials takes place primarily in regions across the Majority World, such as the Democratic Republic of the Congo, Mongolia, and Latin America, where local communities are often displaced, and ecosystems are irreversibly damaged.

Ecological breakdown → Large-scale mining activities often reshape landscapes in profound and lasting ways. The removal of vegetation, disruption of soil layers, and alteration of water flows can lead to deforestation, erosion, and loss of biodiversity—though the degree and type of impact vary widely depending on geography, mining methods, and ecological context. In regions where ecosystems are already under pressure from climate variability or land-use change, mining can further destabilize delicate ecological balances.​

Contamination of surface and groundwater through tailings, chemical runoff, or sedimentation may not always be immediate or visible, but can accumulate over time, affecting both human health and the integrity of more-than-human habitats. Aquatic systems, for instance, are particularly sensitive to changes in pH levels or the introduction of heavy metals. These effects extend beyond local boundaries, with implications for migratory species, food systems, and ecological resilience.​

From a planetary justice perspective, the concern is not only environmental degradation per se, but the uneven distribution of these burdens. Communities whose livelihoods are closely tied to land and water—whether through farming, herding, or cultural practice—may face significant disruptions without adequate safeguards or recourse. At the same time, more-than-human life, while often invisible in policy decisions, bears an equally real cost through habitat loss, toxicity, and altered ecological relations.​

Climate breakdown → AI is frequently positioned as a technological solution to the climate crisis—used in applications ranging from renewable energy optimization to climate modeling. Yet, the upstream activities that make AI possible, including the mining and refining of critical minerals, carry their own carbon footprints. These processes are energy-intensive, often relying on fossil fuel–powered machinery and industrial heat sources that contribute to greenhouse gas emissions.​

The contradiction here is not necessarily in the use of AI itself, but in the systemic disconnect between the promise of digital innovation and the material infrastructures that underpin it. The emissions linked to resource extraction are just one part of a broader chain of climate-relevant impacts—many of which remain poorly tracked or disclosed. This includes the energy used in transporting materials, the loss of carbon-sequestering ecosystems like forests or wetlands due to mining infrastructure, and the cumulative effects of land use change.​

From a planetary justice perspective, the concern lies not only in emissions levels but in who experiences the consequences of climate disruption and who has the capacity to adapt. Human and more-than-human communities located near extractive zones often face compound pressures—from heat, drought, or shifting weather patterns—at the same time that extractive industries alter their local ecologies. As AI development accelerates, a key question becomes whether climate benefits are being realized at the scale and locations where they are most needed—or whether the costs continue to fall disproportionately on those already on the frontlines of environmental change.

Labor injustices → The labor conditions surrounding mineral extraction vary widely depending on region, regulatory oversight, and the scale of operations. Yet across many contexts, extractive industries have been linked to persistent labor challenges—ranging from inadequate health and safety protections to insecure employment arrangements and, in some cases, the presence of child labor. These issues are especially acute in artisanal and small-scale mining (ASM), which often operates outside formal labor frameworks but contributes significantly to global supply chains.​ 

In the cobalt-rich regions of the Democratic Republic of the Congo (DRC), for example, a mix of industrial and artisanal mining sustains the global demand for key materials used in AI hardware. While large firms may adopt formal labor standards, enforcement gaps remain, and informal workers continue to face heightened exposure to physical hazards, respiratory risks, and economic precarity. Reports have documented instances of forced labor and child labor in these mining operations, highlighting the urgent need for improved labor practices and oversight. At the same time, the region’s broader political and economic context—including historical underinvestment, conflict, and governance challenges—shapes the conditions under which labor takes place.​

In South America’s lithium triangle, extraction has triggered tensions not only over land and water but also over the distribution of economic benefits. While mining projects may offer employment opportunities, these are often temporary, specialized, or externally sourced—leaving local communities with limited long-term gain. Moreover, the expansion of extraction in environmentally sensitive areas has, in some cases, deepened existing inequalities, particularly where Indigenous governance structures are not adequately recognized or consulted. Indigenous communities have protested against mining activities that threaten their livelihoods and cultural heritage, calling for more inclusive decision-making processes and equitable benefit-sharing.​

From a planetary justice perspective, labor justice is not just about wages or safety standards—it is about ensuring that those whose work underpins AI’s material foundations are afforded dignity, voice, and agency. This includes supporting community-led decision-making, enabling access to grievance mechanisms, and challenging global supply chain models that rely on labor rendered invisible in dominant narratives of technological progress.​

Resource colonialism → The global supply chains that support AI development are shaped by longstanding patterns of economic and political asymmetry. While many of the minerals essential to AI hardware—such as lithium, cobalt, and rare earth elements—are extracted in the Majority World, the technological value they enable is often realized elsewhere. Ownership of the extraction process, control over pricing, and access to downstream technologies tend to concentrate in countries and corporate centers with greater financial and geopolitical power.​

This uneven distribution is not new. It reflects broader legacies of colonial resource extraction, structural adjustment policies, and commodity-dependent economies that have historically limited the ability of resource-rich countries to pursue their own industrial strategies or set terms favorable to local populations. In many cases, the infrastructure, regulatory frameworks, and financing mechanisms that govern extraction are developed through international partnerships or investor agreements that prioritize global market access over local sovereignty.​

At the same time, it is important to recognize that countries in the Majority World are not passive actors. Some, like Indonesia or Bolivia, have pursued policies aimed at increasing domestic value addition and asserting more control over mineral governance. These efforts, however, often take place within constrained political and economic landscapes, where leverage is limited and environmental concerns may be weighed against promises of development or foreign investment.​

A planetary justice lens invites reflection on how extraction is governed, who participates in those decisions, and how benefits and harms are distributed across time, space, and community. The goal is not just to highlight disparity, but to support alternative models of partnership and accountability—ones that respect ecological limits, uphold collective agency, and challenge the assumption that technological growth must rely on externalizing risk to others.​

The mining of raw materials essential to artificial intelligence–such as lithium, cobalt, nickel, and rare earth elements--is geographically uneven and deeply shaped by historical patterns of resource dependency, trade asymmetries, and geopolitical leverage. While many of these minerals are found in the Majority World, including regions of Africa, South America, and Central Asia, control over their extraction and global pricing often lies elsewhere.

For example, the Democratic Republic of the Congo (DRC) produces over 70% of the world’s cobalt, a critical input for AI infrastructure, but most of it is mined by subsidiaries of foreign companies and then exported for processing abroad. Similarly, South American countries--particularly Chile, Argentina, and Bolivia--host the world’s largest lithium reserves in the so-called “Lithium Triangle”, but much of the ownership and export value is controlled by multinational corporations based in China, the U.S., or Australia. These dependencies reflect not just geology, but also legacies of structural adjustment programs, foreign investment treaties, and commodity-based economies that constrain domestic value-addition and environmental governance.

As global demand for these minerals intensifies--with AI, green energy, and digitalization accelerating parallel booms--resource-rich countries are navigating increasing geopolitical pressure. For some, this has created leverage: Indonesia, for instance, has implemented a nickel export ban to incentivize domestic processing and industrial development. Others, like Bolivia, are pursuing models of state-led lithium governance aimed at retaining more control over extraction and benefit-sharing. However, these strategies are constrained by global market volatility, infrastructure gaps, and external political pressures.

Efforts by the U.S., EU, and other actors to secure supply chains--often framed under the banner of “critical minerals security”--have focused on diversifying sourcing relationships and reducing reliance on any one country or bloc. Initiatives like the Minerals Security Partnership and bilateral agreements with countries such as Zambia and the DRC aim to create alternative corridors for extraction and export. Yet without parallel commitments to environmental protections, labor rights, and community consultation, these partnerships risk perpetuating a model of resource extraction that externalizes harm and reproduces inequality.

A more just geopolitical strategy would not only address where minerals are sourced, but also how governance is shared, whose knowledge counts, and what forms of long-term reciprocity are established between countries, corporations, and communities. Resource-rich nations should be able to negotiate on equal terms and retain agency over the conditions under which extraction takes place--not merely serve as sites of raw material provision in a global AI supply chain.

Viewing raw material extraction through a planetary justice lens involves acknowledging that the physical foundations of artificial intelligence are shaped by choices about how and where materials are sourced, and under what conditions.

A key concern is how extraction decisions are made and whose perspectives are reflected in them. In many cases, affected communities have limited opportunities to influence how land is used, what safeguards are implemented, or how benefits are distributed. This can lead to tensions between local priorities and broader industrial or technological goals, particularly in areas where historical or institutional constraints limit avenues for meaningful consultation.

A planetary justice approach emphasizes the importance of considering these dynamics not only at the level of individual projects but as part of the broader governance of technology development. It encourages practices that recognize the knowledge and interests of communities most directly affected by extraction, and supports mechanisms for inclusive dialogue, environmental oversight, and equitable benefit-sharing.

This perspective also invites a broader ecological awareness—not only treating environments as sites of resource extraction or protection, but as complex systems whose health and integrity shape the long-term viability of technological development. Ecosystems, when altered or degraded, can influence the availability of water, the stability of land, and the resilience of surrounding communities. Paying closer attention to these interdependencies can inform more sustainable, regionally grounded approaches to sourcing and planning.