Though we may not realize it, data centers are an indispensable part of virtually all contemporary composition and rhetorical work. For example, any scholar who studies digital rhetoric or uses their iPhone to take pictures of archival artifacts for later analysis utilizes the power of a data center somewhere in the world. Data centers happen to be so peculiar that they need yet another veiled term, “the cloud”, to obfuscate the true narrative behind how this digital infrastructure works. Behind this supposed cloud lie buildings, rare earths in hardware components, heat exhaust, water consumption, and maintenance staff keeping this infrastructure running (Edwards et al. 431). As an abstraction built into physical technology, “the cloud” is reduced to a screen metaphor that has “no analogy to the affordances engineered into the operating system” as it bears only a faint relation to what it represents (Drucker 216).
Each time your Google Doc saves itself automatically, a saved icon appears as a cloud with a checkmark. “It’s saved!” this icon proclaims. “Where is it saved?” you may ask. It’s saved to a data center, which may exist in The Dalles, Oregon; the Lowcountry region of South Carolina; or Dublin, Ireland, to name a few official Google facilities (Google). We do a lot more online than using Google Docs, though, and data centers power much of it—from hosting a course on Canvas to scrolling on Facebook or Reddit (all of which use Amazon Web Services). As we continue to churn out more writing, images, sound bites, and videos like never before, data centers remain the key infrastructure that undergirds the funny experiences that populate the web, the output of which is only increasing due to AI-generated content. Since the scale of our storage needs seems to be expanding, so is the urgency surrounding the erection of new data centers across the globe (Goldman Sachs).
With this Yack post, we hope to persuade audiences to understand the significance of data centers as rhetorically powerful by analyzing how and why this infrastructure is ubiquitous in the Global North but mostly externally controlled in the Global South. As sites that manage cloud computing, storing data, and powering AI that most internet users require, data centers are consequential in matters of knowledge production and access.
Histories and Geographies of Data Center Infrastructures
In their recent work naming and sketching the parameters of critical data center studies, Edwards et al. direct our attention to the existing scholarship on the location and materiality of data centers and present a framework for understanding data centers relationally, “allowing for deeper engagement with the transnational extractive and logistical infrastructures that support and maintain these [data center]sites” (433). Our work extends the relational assemblage of data centers to consider the frames of the Global North and the Global South, which prompt us to consider how the transnational colonial histories of these parts of the world shape data center usage and how data center usage in turn shapes the information and knowledge production of these places. We interrogate how data centers are implicated in histories of colonial thought, specifically extractivism, surveillance, and ownership, and how these imaginaries are shaping the epistemological terrain of the modern information economy.
Understanding contemporary uses and effects of data center infrastructures requires taking account of their histories and geographies. Scholars who study data centers have documented how current data center infrastructures rely on past infrastructures, creating a kind of palimpsest to examine how pre-existing geographies and ideologies align with these current uses (Edwards, Losh, Johnson, Vonderau). According to Hu, most countries in the Global North built data centers upon infrastructural networks that already existed, particularly sewer lines, railroad tracks, and electricity grids (for this last one, see Vonderau 7). In his book, Hu argues that most of what is considered the cloud rests on “an older structure of sovereign power” (xvi).
These existing infrastructures facilitate the Global North’s continued, expanding investments in Big Tech, but they also reflect histories of colonization and imperialism. By tracing the role of the telegraph in US military histories, Losh presents a rhetorical theorization of infrastructure that prompts us to “ask questions about what aspects of communication are intentionally assumed to be background conditions and consigned to positions of ideological invisibility” (20). Losh argues that telegraphy infrastructure expanded not only the communication possibilities of the telegraph but also US colonization and military imperialism in the Caribbean and in the Pacific, as building these infrastructures required the forced labor of Indigenous and occupied people and the use of these infrastructures furthered the reach of imperial ideology (22-27). The imperial imaginaries that made the far-reaching use of the telegraph possible also explain how other infrastructures displace Indigenous peoples, extract natural resources, and further imperial ideologies (Vonderau, Johnson). Regarding data center infrastructures specifically, Vonderau and Johnson write about Sweden and Iceland, arguing that Big Tech rhetoric relies on implicit imperial rhetoric in order to justify adapting northern landscapes, climates, and energy infrastructures to suit the needs of data centers. Additionally, Vonderau directs our attention to the ways in which even the air itself becomes integrated into data center infrastructures. In these cases, Big Tech companies rely on narratives of embracing the “natural” features of these spaces to explain their continued occupation (Johnson).
While most of those in the Global North seemed to have gained a clear edge based on the colonial and imperialist systems that led to the establishment of infrastructures across Europe and the Americas, specifically, those in the Global South were left scrambling to bridge the digital divide on an uneven playing field. For example, Nigeria, which is in the Global South, currently has 25 data centers in the nation, with 21 in Lagos, 3 in Abuja, and 1 in Kano State, as the country boasts a population of 230 million people. Whereas, Virginia, a state in the US and Global North, has 570 data centers and counting in a location with nearly 9 million people. Although there are complex reasons behind this gross disparity, these numbers send a clear message about where is considered the hub for knowledge production, since those with more robust infrastructure can use and access rhetorical technologies better, leading to the exclusion of an overwhelming number of people.
If the data centers we rely on are built on the infrastructural remains of colonial and imperial ideologies, we must interrogate how these spaces and histories shape our engagement with the newest manifestation of these ideologies. Behind the decisions on how data infrastructure are built lie policies, Big Tech, and local communities that are directly affected by data centers. Although the presence of data centers directly increases technological activities, especially in the Global South by enabling a stronger digital economy, the oversaturation of data centers in the Global North, specifically in the US, enforces surveillance, discrimination, and subpar living conditions in the areas where they are built. Despite the ripple effect of these digital infrastructures across the globe, the oversaturation of data centers in the Global North has been further fuelled by the growing obsession with leading the AI race, as it is considered a key aspect of the next era of global leadership.
In short, we argue that data centers perpetuate the stratification of epistemology, or the material and ideological investment in distinguishing between types of knowledge that can be produced around the world and where this knowledge should be processed and stored. If we define rhetoric as the circulation of power through discourse, or in Dolmage’s words, “the strategic study of the circulation of power through communication” (2), our strategy for studying the significance of data centers requires accounting for the flow of power within these infrastructures–not only the energy and natural resources they need to function but also the rhetorical power that keeps people and companies invested in their use in strategic ways.
Digital Surveillance in the Global North: ICE’s Use of Webloc and Tangles
Mean Worlds Off- and Online
In addition to linking data center infrastructures to colonial and imperial histories, we should also point out that data center infrastructures power what Lievrouw describes as the “mean world” of the internet, which replicates the worst of these ideologies’ social engagements in digital spaces (617). In much the same way as colonizers viewed early colonized territories as lawless, dangerous places where order must be enforced, present-day private and state authorities frame the internet similarly in order to justify surveillance practices (Lievrouw 621). This involves the invasive application of imperialist ideologies that violate individual users’ privacy by gathering, classifying, and storing their data (Lievrouw 621).
For Beck, the surveillance practices within the mean world internet divide one’s online identity into visible and invisible digital identities (126). Invisible digital identities are distinct from the information we willingly share online. For example, a visible digital identity would consist of data that a user makes known about themselves, such as their name and image on their Instagram account. Invisible digital identities are not self-made, per se; instead, they are created by the use of first- and third-party cookies and web beacons, which are embedded into most digital spaces to personalize and target ads (Beck 128). Third parties are given permission to collect personal identifying information about individual users, such as their IP address, age, gender, interests, and browsing behavior (Beck 129-30). This collection of information becomes one’s invisible digital identity—data that we share with third parties without necessarily realizing it, and that provides a scarily accurate version of ourselves as we exist online and in the real world.
Data centers in and of themselves do not collect and store this information, but they are part of the infrastructure needed to do so. The rhetorical power of data centers is such that individual users are often unaware of the ways in which their identities are made known to private and state interests. To return to the metaphor of “the cloud” to describe the seemingly limitless storage capacity of data centers, it’s also worth noting that the infrastructure of the cloud is designed to be opaque, to operate in ways that are inscrutable to the average user. Therefore, the framework of in/visible digital identities gives composition and rhetoric scholars language for unpacking the significance of digital surveillance, a phenomenon becoming increasingly prevalent in recent Immigration and Customs Enforcement (ICE) raids and in the monitoring of ICE protesters across the United States.
Data Extraction and Digital Surveillance
In 2025 and 2026, in the face of increasingly hostile immigration enforcement tactics being practiced on US residents, we also have to contend with personal data being shared with programs like Tangles and Webloc, two digital surveillance tools created by software company PenLink and purchased by ICE (Moini and Finn, Cox, Newman and Burgess, Quintin). These surveillance tools are being used on US residents, whether they are documented or undocumented, criminal or not, and they would not be possible without the computing and storage made possible by data centers (Quintin). And these surveillance practices themselves would not be possible without the histories and geographies of epistemological orientations upon which data centers are built.
Tangles and Webloc work with the infrastructure of data centers, social media, and cellphones to collapse any meaningful distinction between our digital and embodied identities. Tangles works by scraping information that individuals share on social media (visible digital identity), while Webloc works by tracking cellphones in real time if location sharing is turned on (invisible digital identity) (Cox, Quintin). Neither program uses data that would require a warrant to access because it is extracted by mobile data brokers, meaning that any person who uses social media and/or who has a smartphone with location services turned on is susceptible to being monitored by these private interests and their state contractors (Moini and Finn, Quintin). For example, protesters who might have congregated in Minneapolis, Minnesota or Los Angeles, California could be tracked from the location of the protest to their homes, and this location information gathered by Webloc could be supplemented with social media information gathered by Tangles, providing ICE with a significant amount of personal information without even needing a warrant. The invisible digital identities that are created about us make us vulnerable in the real world because the data collected about us is accurate enough to locate us and name us.
Although specific evidence of Tangles and Webloc being used on US residents has not been reported (at least that I [Erin] could find at the time of writing), it is confirmed that ICE has access to these tools and that ICE is monitoring suspected undocumented residents and protesters. In the Minnesota ICE raids of 2025-2026, Minnesota residents participating in “ICE Watch,” or legal observation, reported unknown ICE agents addressing them by name and being told that they were being monitored with facial recognition technology (Frenkel and Krolik). This technology can then be further augmented with social media and cellphone tracking tools to build a body of identification data for each individual who encounters or is in the vicinity of ICE (Frenkel and Krolik; Moini and Elder).
As this small example shows, data centers are a key component in the surveillance of US residents. Data centers make this work possible by storing the data being accessed and powering these tools as they compute. They also facilitate the contracted relationship between private companies like PenLink and state agencies, both of which have much to gain when it comes to extracting personal data. The ubiquity, convenience, and fun that can come with being online elide the steep privacy costs that users may unknowingly be paying, but composition and rhetoric scholars are well-suited to address the rhetorical infrastructures and rhetorical strategies that hide these costs and allow power to continue to be concentrated in private and state interests. As we witness the increasing hostility of ICE agents towards US residents, we must consider how the rhetorical power of data centers impacts vulnerable populations in the Global North.
An Uneven Playing Field for the Global South: A Case Study of Nigeria’s MainOne
Before the efforts made by Funke Opeke, the CEO of MainOne, to build the 7,000km submarine cable running between Nigeria and Portugal, the majority of the nation was running solely on mobile technology led by telecommunication companies such as MTN (Opeke). Before 2010, there were no data centers in Nigeria, and the first fiber optic internet submarine SAT-2 cable had already become obsolete. The direct effect of this infrastructural gap meant that Nigeria’s tech industry was behind its global counterparts, and the necessary capabilities needed to build a digital economy were almost non-existent.
Africa’s Tech Boom
Due to the efforts of MainOne and the infrastructural gap identified by Opeke, Nigeria’s tech economy gained an enabling environment to operate, and the company became a provider of data services to local and international companies, including Google (Carlen, Kazeem). MainOne started as a company accredited for the laying of submarine cables that improved West Africa’s broadband capacity by 15 times and later grew into an ecosystem that includes four data centers across Africa (Equinix).
In 2015, MainOne built its first data center (MDXi) in Lagos, which helped in hosting servers locally instead of relying on Europe for e-commerce and cloud computing. Afterwards, a Tier 3 data center was built in Ghana in 2017, followed in 2018 by another data center in Abidjan. The fourth data center was built in Lekki, Lagos, in 2022 (Delvett). The digital infrastructure in Nigeria then included Undersea cables, data centers and terrestrial cables across Africa to ensure both connectivity and secure channels for cloud computing to bypass public internet (MainOne, Onukwe, Opeke). Online banking, Nollywood streams, content creators, FinTech, and the digital economy were built on MainOne, which is the backbone of what is understood to be the catalyst of Nigeria’s tech revolution.
Existential Crisis of Digital Ownership
Despite the obvious success of MainOne across Africa, global competitors started to enter the market. In 2022, Google’s submarine cable, Equiano, came into the playing field while Meta’s 2Africa was laid in 2023. These cables surpass MainOne by approximately 20 times, as these US-based Big Tech companies had individual capacities of 144Tps (Equiano) and 180Tps (2Africa) as opposed to MainOne with a dwarfed capacity of 8Tps (Nwankwo). The rhetorical effects of the name choice of Google’s undersea cables can be linked back to the transatlantic slave trade since Equiano was a slave abolitionist taken from Nigeria. What was thought to be a memorial to a phenomenal figure might as well be considered an imperialist and extractivist act that undergirds the West’s obsession with Africa’s resources (human and material) and, in this case, Africa’s data.
Due to the presence of competition and the need for “deeper pockets” to prevent MainOne from being obliterated from the market, the company was acquired by another US company, Equinix, in 2022 for $320 million (Nwankwo). Most news outlets tag this transaction as an expansion of Equinix into Africa, since it was the last continent across the globe to have an Equinix owned digital infrastructure (TechCrunch). An article in Forbes even called the buyout a reward for Funke Opeke’s vision to deploy undersea cables in West Africa and build data centers. Despite these narratives across news outlets, what echoes throughout them is the existential crisis that Africa currently has, which makes it almost impossible to own any substantial digital infrastructure.
The largest media platforms, hardware, and software companies across the continent are predominantly owned by Western companies that mostly perpetuate their interests. A single infrastructural project that showcased Africa’s ability to solve its own problems and claim digital agency was quickly overshadowed by yet another Western company under the guise of “expansion.” Similar occurrences have been happening with media companies such as Netflix. Since Nigerians have no say in how Netflix operates, their sudden decision to reduce their investment in African stories showed how fragile a nation can be in the face of externally controlled infrastructure. MainOne, an Equinix company is externally owned, which implies that the data centers and undersea cables that are underlying infrastructure for Nigeria’s connectivity can be controlled, withheld or even manipulated against the interests of the African people. All of this because ownership was transferred both by the owners of company and the African continent whom the infrastructure serves.
Perhaps from a business perspective, selling MainOne to Equinix was an avenue to secure larger funding, ensure competitiveness in the face of stronger competition and avoid the devaluation of the company as a whole. However, this directly addresses the wealth gap between countries in the Global South and those in the Global North, making those in the South perpetual consumers while the North continues to own infrastructure way outside their own terrain. One main concern lies in the inability for the African countries involved to pull out the necessary funding to ensure that ownership is retained. Ultimately, digital infrastructure should be treated not only as an individually owned entity but also with national and continental importance to facilitate data sovereignty.
What is the way forward?
We hope that this Yack article signals to readers why it is important to use rhetorical studies to trace the flow of power when discussing data centers. We’ve explored two examples of data center usage in the Global North and Global South and highlighted the ways in which these facilities enable the flow of power across individual, communal, state, and corporate entities. Data centers store data from localities around the world, but what is stored doesn’t necessarily benefit those in that location. There’s an urgency for the location of data centers, their ownership structures and the aftermath of what they store to be viewed rhetorically and critically beyond the vague narratives that exist online, since there remains a lack of transparency concerning the details of our data.
Works Cited
Beck, Estee. “The Invisible Digital Identity: Assemblages in Digital Networks.” Computers and Composition, vol. 35, 2015, pp.125-40, https://doi.org/10.1016/j.compcom.2015.01.005.
“Companies Using Amazon Web Services (AWS).” Geeks for Geeks, 23 Jul 2025, https://www.geeksforgeeks.org/devops/companies-using-amazon-web-services-aws/. Accessed Jan 2026.
Cox, Joseph. “Inside ICE’s Tool to Monitor Phones in Entire Neighborhoods.” 404 Media, 8 Jan 2026, https://www.404media.co/inside-ices-tool-to-monitor-phones-in-entire-neighborhoods/.
Dolmage, Jay. Disabled Upon Arrival: Eugenics, Immigration, and the Construction of Race and Disability. U of Ohio P, 2018.
Drucker, Johanna. “Reading Interface.” PMLA, vol. 128, no. 1, Jan. 2013, pp. 213–220. Modern Language Association, https://www.jstor.org/stable/23489280
Edwards, Dustin W. “Critical Infrastructure Literacies and/as Ways of Relating in Big Data Ecologies.” Computers and Composition, vol. 61, 2021, https://doi.org/10.1016/j.compcom.2021.102653.
Edwards, Dustin, et al. “The Making of Critical Data Center Studies.” Convergence: The International Journal of Research into New Media Technologies, vol. 31, no. 2, 2025, pp. 429–446.https://doi.org/10.1177/13548565231224157
Frenkel, Sheera, and Aaron Krolik. “How ICE Already Knows Who Minneapolis Protesters Are.” The New York Times, 30 Jan 2026, https://www.nytimes.com/2026/01/30/technology/tech-ice-facial-recognition-palantir.html?searchResultPosition=1.
Funke Opeke, Judith Gardiner Throw Light on Equinix Acquisition of MainOne. YouTube, uploaded by Arise News, 8 Dec. 2021, https://www.youtube.com/watch?v=E86FoQktKSU
Goldman Sachs. “Is There Enough Data Center Capacity for AI?” Goldman Sachs, 11 Dec. 2025, www.goldmansachs.com/insights/articles/is-there-enough-data-center-capacity-for-ai.
Google. “Data Centers: Building for Everyone.” Google. https://datacenters.google/. Accessed Jan 2026.
How Funke Opeke Built and Sold Her Tech Company for $320M. YouTube, uploaded by Delvett, 9 Oct. 2025, https://www.youtube.com/watch?v=2SsTpgup_qI
Hu, Tung-Hui. A Prehistory of the Cloud. MIT Press, 2015.
Johnson, Alix. “Emplacing Data within Imperial Histories: Imagining Iceland as Data Centers’ ‘Natural’ Home.” Culture Machine, vol. 18, 2019, https://culturemachine.net/vol-18-the-nature-of-data-centers/emplacing-data/.
Kene-Okafor, Tage. “Equinix Is Acquiring Nigeria’s MainOne for $320M as It Expands into Africa.” TechCrunch, 7 Dec. 2021, techcrunch.com/2021/12/07/equinix-is-acquiring-nigerias-mainone-for-320m-as-it-expands-into-africa/
Lievrouw, Leah A. “The Next Decade in Internet Time: Ways Ahead for New Media Studies.” Information, Communication & Society, vol. 15 no. 5, 2012, pp.616-38, https://doi.org/10.1080/1369118X.2012.675691.
Losh, Elizabeth. “The Rhetoric of Infrastructure: American Colonialism and the Military Telegraph.” RhetOps: Rhetoric and Information Warfare. U of Pittsburgh P, edited by Jim Ridolfo and William Hart-Davidson, 2019, pp. 19-32, https://muse.jhu.edu/book/68413.
Moini, Nina and Ellen Finn. “How ICE Uses Phone and Internet Data to Identify and Track People.” MPR News, 12 Jan 2026, https://www.mprnews.org/episode/2026/01/12/how-ice-uses-phone-and-internet-data-to-identify-and-track-people.
Moini, Nina and Alanna Elder. “How is ICE Tracking People in Minnesota? An Expert Explains.” MPR News, 8 Dec 2025, https://www.mprnews.org/story/2025/12/08/how-is-ice-tracking-people-in-minnesota-an-expert-explains.
Newman, Lily Hay and Matt Burgess. “ICE Can Now Spy on Every Phone in Your Neighborhood.” Wired, 10 Jan 2026, https://www.wired.com/story/security-news-this-week-ice-can-now-spy-on-every-phone-in-your-neighborhood/.
Nwankwo, J. Emeka. “Redefining Competition and Submarine Cable Ownership: Equiano, MainOne, and the Making of Middlemen.” The Information Society, vol. 41, no. 5, 2025, pp. 279–289, doi:10.1080/01972243.2025.2547676.
Quintin, Cooper. “ICE is Going on a Surveillance Shopping Spree.” Electronic Frontier Foundation, 7 Jan 2026, https://www.eff.org/deeplinks/2026/01/ice-going-surveillance-shopping-spree.
Roscoe, Jules. “The LAPD is Using Controversial Mass Surveillance Tracking Software.” Vice, 29 Nov, 2023, https://www.vice.com/en/article/the-lapd-is-using-controversial-mass-surveillance-tracking-software/.
Vonderau, Asta. “Storing Data, Infrastructuring the Air: Thermocultures of the Cloud” Culture Machine, vol. 18, https://culturemachine.net/vol-18-the-nature-of-data-centers/storing-data/.
