It has been difficult to find research that specifically addresses the concerns of AI data centers from an anthropological perspective. This is no doubt due to the recency of these developments. However, the systems of extractivism and energy politics that fuel projects such as the Monarch Compute Campus are nothing new. In this literature review, we consider the ideologies behind extractivist enterprises, including the unequal distribution of environmental consequences within “sacrifice zones.” Further, we acknowledge the long history of community pushback against extractive projects, particularly in Appalachia. Finally, we consider how larger systems of energy politics and administration made a project like the Monarch Compute Campus possible in the first place. 

Extractive Infrastructure & its Evolution

Liboiron, Tironi, and Calvillo (2018) set the framework for understanding extractive infrastructure by showing that toxicity is not best understood as a result of isolated “bad actors,” but as something shaped through the political and regulatory systems that determine what becomes recognizable as harm. This framework is useful for thinking about the Monarch Compute Campus in terms of impacts such as water use, emissions, energy demand, and land use, which may be interpreted through technical thresholds and narrow standards of evidence that make some harms easier to acknowledge than others. Under this lens, infrastructural change often introduces new forms of harm that are not easily measured or discussed, but remain easy to justify due to an orientation towards innovation. 

Brock and Dunlap (2017) highlight the changing nature of how extractive infrastructure is proposed and justified. With the paper focusing on how the German company (RWE) framed the opening of a new coal mine in the Rhineland region of Germany using tactics such as promoting it as a green project and discrediting activists as stalling progress. This was quite explicitly an extractivist project and as mentioned in Liberon, Tironi, and Calvillio (2018) these development projects, whether coal mines or data centers, are justified through discourses surrounding innovation or how they have become less carbon intensive, hence more green. This source remains relevant on how extractive projects are justified and how they fight back against those opposing them through the soft counter-insurgency of propaganda and hard counter-insurgency of mobilization of state forces.

Energy is often procured through extractive infrastructure. This much is obvious by this point, as clarified by Brock and Dunlap (2017), as well as Liberon, Tironi, and Calvillio (2018). Yet the actual scale of infrastructure required to enable the rapid expansion of the AI industry is difficult to fathom. Rohde et al. (2026) approach the environmental impact of AI through three lenses: environmental justice, climate justice, and multispecies justice. Through these three frameworks, they trace a timeline of environmental theory — however, rather than privileging a single position, they consider the three in tandem as modes of understanding the unprecedented environmental challenges posed by AI. Regarding these challenges, Rohde et al. identify the vast quantities of water and rare earth minerals required to fuel AI technologies. This imperative placed upon AI companies — procuring as many of these resources as possible — incentivizes rapid growth and exploitation. This exploitation is especially prominent in the global south, where these resources rarely benefit the actual communities they come from. 

Like many resources on AI data centers, this paper is incredibly new — which is excellent for our project. It provides some preexisting body of research connecting AI with environmental justice. While Liberon et. al (2018) and Brock and Dunlap (2017) provide an excellent introduction to the politics of extractivism, Rohde et. al (2026) analyze the disparate impact of AI data centers on certain parts of the world. While West Virginia is not a part of the global south, it has a long history of exploitation — and wealth created by West Virginians rarely stays in West Virginia. In addition, this proposed microgrid in Mason county explicitly exists to power AI technologies – and thus, directly relates to the "infrastructure of AI” as discussed by Rohde et al. (2026). 

Nucho (2022) extends this discussion by showing that alternative energy infrastructure emerges not only as a technical response to instability but also as a social and political tool. Her discussion of “post-grid imaginaries” is especially useful because it shows how infrastructural separation, such as microgrids, can be framed as innovation when reliance on civilian grids becomes inconvenient for extractive industries. In the case of the Monarch Compute Campus, the microgrid may be presented as relieving potential strain on local communities, even when the benefits of that infrastructure are not meaningfully available to the public itself as a result.

Sacrifice Zones

On paper, Juskus’s (2023) article provides a brief “history of the sacrifice zone” as a concept — how the term originated and evolved into what it is today. This alone is immensely helpful for understanding modern discussions of climate activism — however, the scope of Juskus’s (2023) analysis goes beyond this. He considers, through the lens of his own ethnographic research, how communities react to the concept of a “sacrifice zone”, and how the term affects activist initiatives. Through comparing ethnographic research and past debates on the concept of the sacrifice zones, Juskus aims to describe: “...how the concept of a sacrifice zone emerged, developed, and was enriched through encounters with diverse populations in various contexts over time…” (Juskus, 2023, 4). In many ways, Appalachia was used as a proof of concept for the sacrifice zone — an example of what rampant and unfettered exploitation could bring the rest of the country. However, he notes that — as spurred on by many indigenous and postcolonial scholars — the sacrifice zone shifted in the public consciousness from an unfortunate inevitability to an inherently unethical product of a broken system. 

The concept of a “sacrifice zone” in general is undoubtedly related to our research project. Historically, West Virginia has often taken on the role of a sacrifice zone, and though this paper does not yet focus on West Virginia, it provides excellent context for more specific case studies. This article is especially relevant, though, as it shows Appalachia as the prototypical sacrifice zone in the United States. It is possible for this to happen again — for Appalachia to become the new frontier for AI infrastructure. Many state officials seem to be hoping for this, at least. As we examine counternarratives to AI as West Virginia’s economic savior, pulling from past scholarship on sacrifice zones (especially in Appalachia) provides an invaluable theoretical backing, for both the remainder of this literature review and our project as a whole. 

William Freudenburg (1992) analyzed how rural communities in the Appalachian region are dependent on extractive industries and the economic and social structures that are impacted as well. In this article, Freudenburg coined the term “addictive economies” to show the evolving relationship between communities and extractive companies. Freudenburg (1992) described the evolving relationship between communities and extractive companies, much like drug addiction; where rural or impoverished communities experience a pleasing “boom” in society when a power plant or coal mine opens nearby. In the beginning, industries in these “addictive economies” create new jobs,increase the population, and bring money into the economy, but over time, long-term consequences or side effects begin to appear; such as the pollution of the environment and individuals, or loss of employment if the resource is exhausted (Freudenburg 1992, 305-307). While Freundenburg (1992) does not directly discuss power plants, his use of this “addictive” metaphor clearly shows the negative impacts of extractive industries in a straightforward format. This article provides a different perspective to the debate on the economy versus the environment, than Kojola (2017) by exploring the role of industry dependence on rural communities. Similar to McGarity’s (2019) work, however, Freundenburg (1992) addresses the importance of possible job opportunities and economic stability that large industries tend to idealize. 

Fredricksen and Himley (2019) offer an additional theoretical lens to analyze sacrifice zones positing the concept of extractive frontier. They define this term as zones where people are alienated from resources as corporate entities seize them. The important part of this paper that connects to the other sources is the fact this is an analysis of the strategies that “that firms and their allies deploy to secure and preserve the transformed relations of land and resource access upon which accumulation relies” which is relevant when talking about the history of the coal industry and other extractive industries within West Virginia. This transformation of relations of land and resource seems to be the basis of sacrifice zones as after these land relations are transformed there are zones where extraction without care occurs. This is very relevant in terms of the discourse surrounding AI Data centers where activists and researchers are raising alarms about the resource usage of the projects which could potentially set in motion multiple decades of extraction on those plots of lands. (Rohde et al. 2026)

Nading (2020) adds an important anthropological layer to this discussion by showing that toxic exposure is not just a singular event or an isolated bodily condition, but an ongoing reality shaped by harmful substances, the infrastructures that distribute them, the institutions that manage or define them, and the everyday conditions through which people experience them. This is useful for the literature on sacrifice zones because it shifts attention away from toxicity as something recognized only through formal measurement and toward toxicity as something lived, interpreted, and navigated through uneven systems of responsibility and trust. In this sense, Nading helps show that sacrifice zones are not only produced through large-scale extractive development, but also through the everyday conditions in which communities come to understand harm, especially when official channels of reassurance or accountability are experienced as partial, delayed, or compromised. His more-than-human approach is also important because it treats community interpretations of industry and risk as meaningful forms of knowledge rather than as secondary to technical assessment. 

Nicholas Stump, Head of Research and Reference Services for the WVU College of Law, established early on in his book Remaking Appalachia: Ecosocialism, Ecofeminism, and Law, that the Appalachian region has historically been an “energy sacrifice zone”. In this zone, extractive industries, like coal and natural gas, have destroyed the environment and very nearly the economy, and with poor environmental law, will continue to do so (Stump 2021, 1). Based on previous research, and as the title points out, Stump (2021) theorizes that if systems were ever able to restructure in an ecofeminist and ecosocialist manner, that is very likely to increase the material-based quality of life for many in Appalachia. This large change could lead to a major cultural shift in Western society, where there would be more focus to meet basic needs over hyper-consumerism (Stump 2021, 4-6). One of the most important aspects of this major reform comes from West Virginia’s long history of grassroots activism. Stump points out the significance of the Mine Wars, as well as the more recent teacher strike, in combatting “the system” across race, gender, and class (Stump 2021, 7; 165). Grassroots activism has continuously played a major role in West Virginia’s history, so turning attention to environmental issues might inspire another transformation in the Appalachian region or even globally Similar to McGarity (2019), Stump (2021) refers heavily to environmental law, and West Virginia’s long history with poor legislation on pollutants, and bridges these issues to the need for activism.

Community Response & Activism 

Renfrew and Pearson (2021) help frame the literature on community response to petrochemicals and activism by distinguishing between pollution legacies and toxic events. Rather than treating contamination as politically meaningful the moment it exists, they show that harm becomes socially legible through processes of discovery, narration, and contestation. Their concept of a “toxic continuum” is especially useful because it frames response not as a simple shift from awareness to resistance, but as a movement across invisibility, suffering, resignation, and refusal. This perspective suggests that environmental harm does not automatically produce collective action; instead, response depends in part on whether harm can be recognized, named, and attributed within broader struggles over responsibility and accountability.

The advent of generative AI is a relatively new development — but as Bell and York (2010) reiterate, the extractivist ideology associated with it has existed for decades. Using a neo-marxist lens, they examine how extractive industries manufacture legitimacy among communities; this includes communities they are actively harming. Through analyzing the economic (and eventually, ideological) dependence of communities in West Virginia on these extractive industries, Bell and York illustrate how: “many of the people suffering the most acute costs of ecological degradation are some of the least likely to fight against the treadmill processes because of their economic dependence on the destructive industries” (Bell and York, 2010, 115). Coal companies, for instance, actively associate themselves with visions of upward mobility and rural masculinities, both of which are often attractive to working-class men. Bell and York’s framing emphasizes the connection between economic growth and environmental degradation — with the important note that the “treadmill of production” often leads to fewer jobs, not more (Bell and York, 2010, 112). This fact is intentionally obfuscated by these companies, resulting in a general communal attachment to these extractivist industries — in particular, the coal industry. 

While this article may, at least at first glance, appear to only focus on the coal industry, it establishes a crucial background for understanding community identities (and thus, community movements) in West Virginia. When approaching social movements, it is important to understand what might prevent a social movement from manifesting — and Bell and York do an excellent job of this. Since AI-fueled industrial development is so new, looking back at community reactions to coal provides a barometer for recognizing common patterns. While West Virginia was previously defined by its relationship to coal, officials such as Jim Justice now appear poised to welcome in generative-AI as a new major export — and it is crucial to draw the connection between these two threads. 

Similar to Bell and York (2010), Erik Kojola (2017) examined the beliefs and cultural identities surrounding the works in the resource extraction industry and their relation to environmental movements. While focused primarily on the Keystone XL Pipeline, a proposed expansion designed to transport oil from Alberta, Canada to the Gulf Coast of the United States, Kojola (2017) also investigated the connection between the environment, labor, and capitalism in Appalachia; further examining the “job versus the environment” debate. The communities and workers who rely directly on resource extraction, whether for work or capital, are directly faced with both environmental and job-related hazards that come with the polluting industries (Kojola 2017, 896-897). Like Freudenburg (1992), Kojola (2017),does not attribute communities’ dependence on the resource extraction industry to only capitalist ideology, instead, he also examined how environmental groups have historically neglected economic and class issues, which impact the daily life of workers. Kojola also explored the ideologies and identities associated with resource extraction, including the symbolism of masculinity, whiteness, nationalism, and progress; and how they appear in the media (Kojola 2017, 897-898). One major proposed benefit from the microgrid industry is economic stability and job security. In West Virginia, a state composed of loyal hard workers and faithful providers, job security is crucial for survival, regardless of the work’s impact on the environment or self. Shepherd-Powell (2025) also researched job security and economic benefits seen in West Virginia, as the coal industry shifts to AI. Kojola’s (2017) research challenges this debate of economy versus environment, by determining the two areas can coexist. 

As has been well-established, coal is everywhere in rural Appalachia. This remains true even today, a fact which Shepherd-Powell uses to anchor her ethnography of Wise County, Virginia. The bulk of her ethnographic work takes place in 2011 — just before the 2012 presidential election. However, her political questions are grounded in the economic and political realities of living in the aftermath of coal. For many rural Appalachians, the booming days of coal are a reality that they desperately want to return to — and as Shepherd-Powell writes, this is fundamentally because: “It was a good place to live because people had money and the community was thriving.” (Shepherd-Powell, 2025, 7) Considering the rates of poverty in Appalachia, the overwhelming community desire for jobs often dominates political discourse. Shepherd-Powell uses her ethnography to explore the unease that permeates the legacy of coal — what does it mean to be a “friend of miners” vs. a “friend of coal” (Shepherd-Powell, 2025, 5)? How do communities cope with the constantly shifting policy around coal? And how do political currents in Appalachia reflect discontent with voters’ current realities?

Much of Shepherd-Powell’s work here has a political focus — and, notably, it pulls from ethnographic fieldwork done over a decade ago. However, even a cursory look at modern West Virginian politics will reveal uncannily similar sentiments : coal was the key to past prosperity, and there remains a way to claw said prosperity back. In the case of legislation around AI, the narrative is largely the same — according to some politicians, AI will be the new coal, powered by its intermediary, natural gas. Yet it remains to be seen how Appalachian people will respond to this — can this nostalgia for coal successfully be shifted to a far newer, far less understood technology? As we begin our fieldwork in Mason County, Shepherd-Powell’s work provides an invaluable compass for understanding the current political and economic state of West Virginia through an anthropological lens. While Bell and York’s (2010) and Kojola’s (2017) papers are relatively older, they can be combined with Shepherd-Powell’s work to create a deeper knowledge of extractivism and community responses, particularly in West Virginia. By using this compass, we hope to understand the future of a new extractive industry in West Virginia.

Energy Politics

Dominic Boyer (2019) introduced the term energopower. Energopower is taking the framework of bipower, mechanisms used by the state uses to control the lives of people, developed by French philosopher Michel Foucault and applying it to the diffuse power relations in energy politics, specifically  exploring“ how energie forces and infrastructures interrelate with institutions and ideations of political power.” (Boyer 2014)  This analysis offers a useful theoretical basis as energy politics pervades the literature related to extractive industries as most are linked towards the energy needs of a community whether that be a nation or a town which obviously intersects political and commercial interests. 

Another key aspect to consider in the discussion of future microgrids and the growing demand of artificial intelligence is what resources will be used to run this infrastructure. McGarity’s (2019) research provides an in-depth analysis and exploration of the power plant industry, legislation, pollutants, and resource extraction. By also examining recent events from both the Obama and Trump Administrations, McGarity expands heavily on the political aspects of the power industry, and how it has evolved over recent times (McGarity 2019, 206-220; 321-324). As of recently, politics have continuously alternated between environmental standpoints and legislation; creating consistent instability across the nation, especially in Appalachia. McGarity proposed numerous solutions to fix both regulation and business issues in the power industry, including “decoupling” charged rates of electricity, cybersecurity, and encouraging demand reduction (McGarity 2019, 331-335). The proposed microgrid in Mason County, West Virginia does not explicitly state what resources will be used to power the plant, but most plants have switched to natural gas. McGarity (2019) goes into great detail when discussing how power plants work, what they produce in terms of resources and pollutants, and why they are used; providing a clear picture of possibilities before, during, and after the Mason County microgrid opens. 

Dunlap (2017) highlights one of the most paramount examples of energy politics or energopower in motion would be the construction of the Bíi Hioxo Wind Park in Juchitán de Zaragoza, Oaxaca, Mexico that was built in 2014.  There was a confluence of ideologies and powers on the ground shaping the discourse on the ground.  There were (mostly indigenous) activists and the Mexican government (local and national) who aimed to shape discourses around this project. The government and corporations argued this project would be useful for the green transition while activists argued that this project was not required to provide power to the public and was only beholden to its shareholders who ironically are made up of high carbon emitting corporations. In the most unadulterated way, energopower can be termed the harnessing or management of electricity and fuel in cases like this overlapping with biopower where these questions of electricity impact how people live. This example also showed this isn’t solely about discourse but also about force with security forces and state forces targeting. An example of more historical discourse tinged with energopower is the discourses regarding the mass building of power plants during the Second World War Two towards discourses of regulation during Clean Air Act towards modern regulations that vary between Democratic and Republican administrations. (McGarity 2019).

Vonderau (2018) adds an important dimension to the literature on energy politics by showing that data centers are not introduced simply as technical infrastructure, but as development projects made attractive through what she calls “technologies of imagination.” In her ethnographic work on Facebook’s first European data center in Luleå, Sweden, she shows how planned visibility, cultivated invisibility, and future-oriented narratives help frame the “cloud” as a transition into a cleaner and more modern future. Her contribution is especially useful because it makes clear that the politics of digital infrastructure are not limited to energy use or land use alone, but also include the ways corporations and municipalities shape what can be known, what remains uncertain, and what comes to feel inevitable. Read alongside Boyer, Vonderau helps show that energy politics are not only about the management of fuel, electricity, and land, but also about the production of narratives that make new infrastructures appear necessary, modern, and politically acceptable.

The Appalachian region has had a long history with extractive industries and the consequences that follow. In their article “Environment, Health, and Justice: TRACING THE CONNECTIONS IN GLOBAL MOUNTAIN REGIONS”, Anglin, Button, and Molina-Rosales (2018) examined different case studies surrounding the negative impacts of extraction, specifically in the Appalachian region, including Logan County, West Virginia and Warren County, North Carolina (165;168-170). By first examining previous federal action, or the lack thereof, a clear image is painted of the lack of effort and environmental legislation in the Appalachian region since (Anglin, Button, and Molina-Rosales 2018, 164). Recently, anthropologists and other researchers have targeted poor quality of living near extractive industries, by not only examining a decrease of crops and wildlife, but also other federal policies that maintain “quick fix” approaches to disaster (Anglin, Button, and Molina-Rosales 2018, 168-171). For many living in Appalachia, extractive industries are a part of everyday life, often going unnoticed. While some residents seem to be upset with the proposed microgrid in Mason County, West Virginia, many are unaware of the situation as it has been kept relatively quiet and out of the public eye. One of the most transformative approaches to this environment debate is the emergence of grassroots activism, by bringing attention to real-world situations. In doing so, people living in Appalachia become aware of the consequences of extractive industries, which are almost always overlooked on a federal level.

One of the major strains of discourse in energy politics would be the idea of regulatory capture or regulatory neglect. This is highlighted in Chanell (2011) which goes through how environmental laws were ignored or watered down by corporations like Massey Energy and regulatory agencies such as the Mine Safety and Health Administration (MSHA). It showcases how these discourses surrounding energopower are literally life and death, highlighting the Upper Big Branch Mine Explosion where 29 miners were killed in a mine that was not abiding by regulations. This is paramount in exploring as current discourses surrounding how to regulate power plants (McGarity 2019) and AI data centers are at the forefront of politics with bills such as “the One Big Beautiful Bill” originally aiming to halt any regulation of AI at a state level. As Vonduru (2018) posits these ideas of future oriented narratives shape these changes in regulatory structures as inevitable. While they are talking about the creation of data centers for cloud computing, this seems to historically apply to projects of extraction in other zones such as coal or the current discourses surrounding AI data centers.

These major themes found in our literature review point to a system deeply mired in both past extractive industries and imagined futures of energy production. As this background informs our anthropological understanding of AI data centers and the response they evoke, these concepts can be applied to the specific circumstances of those living in Mason County, WV. The literature featured above fueled the questions we prepared as we began our in-person ethnographic research. 

Anglin, Mary K., Gregory V. Button, and Dolores Molina-Rosales. 2018. “Environment, Health, and Justice: TRACING THE CONNECTIONS IN GLOBAL MOUNTAIN REGIONS." Global Mountain Regions: Conversations toward the Future. Indiana University Press. 164-76.

Bell, S. E., & York, R. (2010). Community Economic Identity: The Coal Industry and Ideology Construction in West Virginia. Rural Sociology 

75(1), 111–143.       

Boyer, Dominic. 2014. “Energopower: An Introduction.”  Anthropological Quarterly 87, no. 2: 309–33. 

Brock, Andrea, and Alexander Dunlap. 2017. “Normalising Corporate Counterinsurgency: Engineering Consent, Managing Resistance 

and Greening Destruction around the Hambach Coal Mine and Beyond.” Political Geography 62 (October): 33–47. 

Channell, Emily S. 2011. “Coal Miners’ Slaughter: Corporate Power, Questionable Laws, and Impunity.” North American Dialogue 14 (1): 

12–22. 

Dunlap, Alexander. 2017. “Counterinsurgency for Wind Energy: The Bíi Hioxo Wind Park in Juchitán, Mexico.” The Journal of Peasant 

Studies 45 (3): 630–52. 

Frederiksen, Tomas, and Matthew Himley. 2019. “Tactics of Dispossession: Access, Power, and Subjectivity at the Extractive Frontier.” 

Transactions of the Institute of British Geographers 45 (1): 50–64. 

Freudenburg, William R. 1992. “Addictive Economies: Extractive Industries and Vulnerable Localities in a Changing World Economy.” 

Rural Sociology 57 (September): 305–32. 

Juskus, Ryan. 2023. "Sacrifice Zones: A Genealogy and Analysis of an Environmental Justice Concept." Environmental Humanities 15 (1): 

3–26. 

Kojola, Erik. 2017. "(Re)Constructing the Pipeline: Workers, Environmentalists and Ideology in Media Coverage of the Keystone XL 

Pipeline," Critical Sociology 43 (6): 893-917.

Liboiron, Max, Manuel Tironi, and Nerea Calvillo. “Toxic Politics: Acting in a Permanently Polluted World.” Social Studies of Science 48, 

no. 3 (2018): 331–349.

McGarity, Thomas O. 2019. Pollution, Politics, and Power: The Struggle for Sustainable Electricity. Harvard University Press. 

Nading, Alex M. “Living in a Toxic World.” Annual Review of Anthropology 49 (2020): 209–224.

Nucho, Joanne Randa. “Post-grid Imaginaries: Electricity, Generators, and the Future of Energy.” Public Culture 34, no. 2 (97) (2022): 265–

290.

Renfrew, Daniel, and Thomas W. Pearson. “The Social Life of the ‘Forever Chemical’: PFAS Pollution Legacies and Toxic Events.” 

Environment and Society: Advances in Research 12, no. 1 (2021): 146–163.

Rohde, Friederike, Zhyar Nasruddin, Elizaveta Kotova, and Sabine Ammon. 2026. “Navigating Justice in AI Lifecycles: Ethical Perspectives 

on Infrastructural Prerequisites and Their Environmental Impacts.” AI and Ethics 6 (1): 155-174.

Shepherd-Powell, J. (2025), “We're Not Going to Be a Coal Superpower Anymore”: Political Discourse, Elections, and Everyday Economics 

in the Coalfields of Southwest Virginia. Journal for the Anthropology of North America, 28: 1-11. 

Stump, Nicholas F. 2021. Remaking Appalachia: Ecosocialism, Ecofeminism, and Law. Morgantown: West Virginia University Press. 

Vonderau, Asta. “Technologies of Imagination: Locating the Cloud in Sweden’s North.” Imaginations: Journal of Cross-Cultural Image 

Studies 8, no. 2 (2018): 8–21