Digital Bandung
In recent years, it has become common—even unavoidable—to refer to the internet, AI, and digital capitalism in terms of empire and colonialism. A partial list of recent high-profile books that use the language includes Empire of AI (2025) by Karen Hao, Silicon Empires (2025) by Nick Srnicek, The New Empire of AI (2025) by Rachel Adams, Digital Empires (2023) by Anu Bradford, and Cloud Empires (2022) by Vili Lehdonvirta. These join influential work by Nick Couldry and Ulises Mejias in their paired books, The Costs of Connection (2019) and Data Grab (2023), and their 2018 article, “Data Colonialism,” which has over 2,000 citations on Google Scholar (note for non-academics: this is a lot). Honorable mention goes to an article on “digital colonialism” by the South African scholar Michael Kwet, which tops 1,000 citations. Hardly a week goes by without a think piece with a title like “Big Tech is Building Empires,” as the economist Grace Blakeley wrote recently, or “AI is in its empire era,” as the STS scholar Kate Crawford posted last year.
The turn to empire is notable first for displacing a previously preferred language of mind control. After Edward Snowden’s revelations in 2013 about Silicon Valley’s complicity in mass surveillance triggered the first phase of the so-called techlash, critics gravitated to language of behaviorism and anxieties about brainwashing from the mid-twentieth century. Our internet-connected phones and laptops had put us in Skinner boxes; we were the pigeons pecking at the pellets according to a pattern—at the mercy of “attention merchants” in “the shallows.” “The technology that connects us also controls us,” read the tagline for the influential Netflix documentary The Social Dilemma. Shoshana Zuboff—who had taken classes with B.F. Skinner himself as a youngster at Harvard—warned in her 2019 bestseller The Age of Surveillance Capitalism that Silicon Valley now had “the instruments and methods that can impose Skinner’s technology of behavior across the varied domains of everyday life right down to our depths.”
The shift from psychology to political geography is a good one. It centers questions of power and authority alongside issues of land and energy. The reliance of citizens, states, and pension funds on the financial fates of a small number of US tech companies is a startling development of the recent past. It is sensible to search for a term proportionate to our astonishment—empire talk is a rhetoric of shock.
Yet invocations of empire and colonialism also risk being trapped in the register of polemic rather than analysis. As with the category of “techno-feudalism” effectively critiqued by Evgeny Morozov, it is not always clarifying to explain something new with reference to something old—especially if doing so relies on a two-dimensional view of the past. Actual scholarship teaches us that using the word “empire” as a straightforward synonym for a systematic, centralized, and overwhelming form of rule can be precisely wrong. Turning to the literature on empires for more applicable insights can help sharpen our understanding, our language, and thus, our critique. In what follows, I first suggest some of the shortcomings of the current usage of the category of empire and propose three other terms that have been ignored or underplayed in the scholarship on data colonialism: the comprador, the creole, and counter-colonization.
The Empire of the Megacorp
In the classic definition, an empire is a large expansionist polity comprising diverse populations ruled through a politics of difference, hierarchy, and nested sovereignty. Most historians consider it an extinct form. Its death knell came at the end of World War II, as a new norm of national self-determination was championed by the world’s two superpowers, the United States and the Soviet Union, and institutionalized in the United Nations. One by one, and sometimes in bunches—sometimes by peaceful negotiation and at other times through bloody conflict—those parts of the world governed by European imperial powers found their own colors and their own names, their own flags and their own seats in the UN General Assembly. (The seventeen islands and archipelagoes that remain “non-self-governing territories” are the exception that proves the rule.)
To many, however, the apparent end of empire hid a paradox. The very nation that posed as an advocate for a world of nations turned into an empire of its own. This sparked an ongoing debate about whether the US remained an empire even after the “greater United States” that stretched from the Philippines to Cuba was whittled down to Puerto Rico and scattered Pacific atolls. Does empire have to mean direct territorial administration? Or can it be performed more indirectly through the power of the dollar as the global reserve currency, for example, or through locked-in power in the world’s international financial institutions? Or more directly still through an “empire of bases,” which can become staging grounds for so-called police actions to support allies and dislodge adversaries and enemies?
One thing seemed clear in this discussion: Empires were a form of statecraft. This was true even though in the first phase of imperial conquest, in the era of the so-called merchant empires, the power in question—from the Dutch East Indies to New France—was a corporation working with a royal charter. But that era of the “company-state” or what the historian Philip Stern calls “venture colonialism” was seen by most scholars as a superseded form, especially by the era of so-called high imperialism in the late nineteenth century. Privateers and freebooters were often the advance guard, but the flag followed, and imperialism for the most part was a state project.
The US exception was another discourse that rumbled through the Cold War decades, less in the discipline of history as such than in broader political discussions and the social sciences: the assertion that private companies had become their own kind of empires. From “coca-colonization” in Western Europe to Firestone’s “empire of rubber” in Liberia to the speckling of the globe with Export Processing Zones, transnational corporations seemed to exercise a de-territorialized form of power, coordinating global supply chains through internal forms of planning that made them seem almost state-like. This vision became fodder for the still-influential imaginaries of cyberpunk and anime, in which conventional states and governments are all but eclipsed by the power of what has come to be labeled the “megacorp.” Think Weyland-Yutani in the Alien franchise or the Tyrell Corporation in Blade Runner.
Some of the current crop of works on digital empires follow this line of interpretation: The megacorp has slipped the bonds of territorial states to extend its power globally. In a world where far-flung customers and contractors log on to Open Desk, Mechanical Turk, Netflix, and Amazon Web Services to communicate, label data, moderate and stream content, or store their research, the conundrums of jurisdiction, oversight, and regulation multiply.
As the slightly musty cyberpunk reference implies, however, these dynamics are not entirely new: They are familiar from the era of the transnational corporation when poorer nations had to worry that their policy choices would trigger disinvestment—or proxy military intervention—by US-based firms.
Jack Goldsmith and Tim Wu argued twenty years ago in their epochal book, Who Controls the Internet?, that we should not allow the mystifications of digital rhetoric to convince us that old-fashioned states had somehow been overcome. In the decades since they wrote, the ability of states to censor or block digital services—or coerce private providers to conform to state mandates—has become ever more pervasive. Few today would buy the fairy tales about the borderless Internet that enchanted cyber-libertarian pioneers in the 1990s and mainstream boosters like Thomas Friedman in the 2000s.
The dependency of governments on the capacities provided through subscription software has since deepened, but dubbing such private-public tensions “empire” does not sharpen interpretation. Indeed, as Anu Bradford observes in her book, we have long moved on from the idea that digital firms operate unfettered by territorial power. The “digital empires” she describes in the EU, the US, and China already are fusions of government and corporate power. What she calls “empires” are really varieties of digital capitalism, institutional arrangements with varying levels of state direction of investment, and regulation. Putting an “empire” label on the problem of transnational corporate power does no harm, but it also adds little analytically.
The Empire of Emotion
What, then, is the attraction of empire talk for tech critics? For a world of adults raised on Star Wars, it carries an emotional payload with overtones of a particularly merciless form of exploitation and evil. Consider how some of the books are attracted to a stark moment of early modern imperial history. During the conquest of the Americas and the establishment of the Viceroyalty of New Spain, indigenous populations were not simply wiped out, although many died through warfare and vastly more through epidemic disease. The surviving population was drawn into a system of tribute and coerced labor in agriculture, mining, and other forms of production, generating wealth for local colonial authorities for the Spanish Crown to fund its wider imperial wars.
The practices of colonial administration included formalities that seem absurd today. These included rituals of possession, with new arrivals planting a cross or pronouncing an edict so as to claim dominion over an area, at least in their minds. In some encounters with indigenous communities, this came in the form of the so-called Requerimiento: several paragraphs explaining to the now-subjected population that they could accept rule under the Spanish and the superiority of the Catholic Church or face dispossession and enslavement.
The painful punchline of this historical moment is that those to whom the terms were being dictated understood not a word of what they were being told, and thus any impression of consent or surrender was entirely fanciful. To more than one contemporary author, this moment of encounter resonates with the actions we take as web browsers and consumers when we accept, through a single click, lengthy terms and conditions without the wherewithal or legal fluency to understand what we are accepting.
Couldry and Mejias, who use this example in two of their books, argue that the moment captures perfectly the practice of data colonialism: They see it as the “appropriation” of information, activities, behavior, and habits by data colonizers on formally approved grounds that are nevertheless asymmetrically weighted against the users. They argue that such moments occur chronically throughout our interactions with our devices. Through our actions, we are creating what they call “data territories,” which, through their transformation into exchangeable property, become data colonies in which we come to live a kind of doubled existence, even if we do not always realize that.
Here the straining of the analogy is already clear. For those living under direct colonialism, subordination is very difficult to ignore. The harvesting of personal data is different: it is a form of oppression that remains largely invisible to those it affects, and that we often need a third party to reveal to us. To make the analogy with colonialism hold, one would need to invoke the contested notion of internalized oppression.
Scholars who use the language of data colonialism and data empire argue that we must be attentive to those moments where our subordination becomes visible. Examples include manipulative advertising during momentous campaigns, such as Brexit and the 2016 US presidential election, and, for those whose livelihood depends on gig work mediated through an algorithm, the unpredictability of losing a ride or failing a job interview because of an AI-enabled assessment tool. Since the rollout of AI targeting tools like the Maven Smart System, the existential nature of these dangers has become even easier to notice. These are moments when we suddenly see the empire that is always there behind the tempered glass of our screens.
Even if it contributes little analytically, the invocation of “empire” can still function as a tool of mobilization. It can help produce a more active public awareness of algorithmic discrimination and other forms of digital subjection. In some cases, these are indirect and opaque. In others, they are more tangible: living beside polluting data centers powered by gas generators, or enduring the extraction of cobalt from one’s home province. If “empire” is a word that channels political outrage against such injustices, then why not use it?
Digital Compradors
Yet if empire works as an emotional category, the larger problems also start here. For many scholars, sustaining this analytical frame requires a stark contrast between them, the colonizers, and us, the colonized. Dominated by images of the Roman phalanx, the steel storm of the Reich, or Darth Vader’s bridge, empire often connotes the centralization of power and seamless domination. But this better describes empire’s successor, the nation. Empire was defined by delegation, nested sovereignty, and rule through local authorities—figuring out minimal outlays for maximal returns, not smothering in excess.
Many authors endorse a dynamic in which the plundering colonizer extracts fortunes at the barrel of a gun and leaves nothing but poverty and death in its wake. But empire itself did not usually work in this clear-cut way. It rather operated through a middleman called a comprador class.
Couldry and Mejias mention what they call the go-between, referring to indigenous populations that were willing to act as translators or interpreters for fear of worse consequences. But this was not a fleeting moment of transition at the beginning of plunder. The entire history of colonialism was organized around co-opting local populations, either as economic actors who were able to source the goods that the colonizers were interested in or as literal governing populations, as in the practice of indirect rule that characterized many European empires.
Colonialism was not simply a one-way practice of extraction, to use a favorite term of the authors in this subfield; it was also, however asymmetrically, a process of exchange. Empire sought out goods that were desirable to local populations because of their scarcity or technological advancement, and in turn it took from colonial populations goods—whether spices, beaver pelts, copper, or human beings as chattel—that were of interest to the colonizer. These trade objects were used to expand forms of exchange that were not always directly imperial. Consider the fact that the Potosí silver mine in Bolivia was important not primarily for the silver it sent back to Spain, but because silver was a good of interest to the Chinese empire, which still adhered to a silver standard and had little interest in many other European commodities. Empire did not just take. It also connected—leaving blood streaks in its wake.
Many products of imperial trade were used to enrich the coffers of the imperial power, but they were also used to enhance productivity, as in the case of sugar, tea, and coffee sent back to workers in England and France. At the same time these products expanded the sphere of pleasure and variety of consumption that the colonizing population could fold into everyday life and expectations.
The payoff of pleasure is worth mentioning because the literature on data colonialism often ignores it. In a recent book, Kieran Healy and Marion Fourcade write about what they call the “box of delights” that our phones and other devices represent. We subject ourselves to the requirements of our new “tech lords” not simply because we do not understand their language and have no other choice, but because we are given real pleasures in return: the hedonic experience of drifting along streaming videos, absorbing for hours the whispers of people we admire, and avoiding burdensome trips to distant retail outlets, which grants us time for more meaningful forms of human interaction.
These are all certainly part of what Couldry and Mejias call “the myth of convenience.” These pleasures are not only myths, and it makes little sense to try to understand the onward march of digital empires without seeing them as purveyors of a never-ending bliss—in the sense not only of addiction, but also of genuine access to some of humanity’s kaleidoscopic creativity. People who denounce digital empires are often left facing this paradox: What if people don’t want to be liberated?
Digital Creoles
The language used in the digital-empire literature is that of “us versus them”: us the users, them the owners. We, the users, seem to be working double shifts, both for our actual employers (or ourselves) and also for the Mark Zuckerbergs and Larry Ellisons at a distance, as all our activities are mediated through the services they own directly, the data we unwittingly send their way, or the backend of cloud services that keep the whole thing functioning.
But does it make sense to think of users of apps as nothing more than digital indigenous populations being expropriated? The answer—so obvious that it is strikingly absent from much of the literature on digital empire—is no. Big Tech companies constitute some of the most important components of the US stock market, and more than half of the American population is plugged into them, whether directly through stock ownership or indirectly through pensions, mutual funds, and retirement accounts.
The move from the 1970s, when most stocks were owned by individuals, to the present, when most are owned by so-called institutional investors, has been called asset-manager capitalism. Others have suggested more evocatively that it resembles a kind of pension socialism or universal ownership. A recent book on the topic bears the striking title Our Lives in Their Portfolios. Yet a more accurate description would be “our lives in our portfolios,” since the appreciation of assets within the broader stock market is directly trackable on the dashboard of one’s Fidelity or Vanguard account regardless of one’s level of financial literacy.
The people who profit from the harvesting of data exhaust—and what the digital empire literature calls the creation of “data colonies”—are, in fact, the very users themselves. The compensation packages of CEOs may mock any sense of social justice, and the use of dual-class share structures may make it difficult to budge their internal practices. But this does not change the fact that we, in the Global North, are in an Ouroboros-like relationship with our tech companies that is hard to find a parallel for in the history of empires.
The beneficiaries of digital empires are not merely the conquistadors with cutlasses; they are the many people whose financial futures are determined by the performance of tech stocks in the market. It is common to refer to users as “digital natives.” “We are the native peoples now whose tacit claims to self-determination have vanished from the maps of our own experience,” writes Zuboff. But settler populations who intermarried with indigenous populations like the French métis were described by a different word. They were “creoles.”
What would it mean to think of ourselves not as “digital natives,” with all the unsavoriness of that phrase, but more as “digital creoles”? We are often both foreign and at-home, exploited and beneficiaries—shaped by our relationship to our devices, with our financial fortunes bound up in them. What politics might follow? In an influential argument, Benedict Anderson made the case for what he called “creole nationalism” that emerged in Latin America as print capitalism made possible a sense of shared identity and “imagined communities” distinct from the metropole. He sees this as the beginning of the modern nation-state. Aris Komporozos-Athanasiou has updated Anderson’s thesis to ask whether digital technologies are allowing for what he calls “speculative communities” or what were formerly called “netizens.” In this framing, digital capitalism is not a one-way dynamic of exploitation but an oppressive system that nonetheless produces—even in spite of itself—the possibility for new forms of collective political understanding.
Now that the promise of app-driven social movements has faded, it seems mawkish to dream of “Twitter Revolutions” and “Facebook Revolutions” as people did fifteen years ago. But what were the #MeToo and Black Lives Matter movements but similar dynamics brought into the streets of the superpower? Though it wasn’t described that way, Joe Biden’s election in 2020 was a Facebook Revolution of sorts, and it was preceded by the largest street protests in US history, which so plagued the Silicon Valley “emperors” (as a forthcoming book by Cecilia Rikap refers to them) that Elon Musk set about redesigning social media platforms so that nothing of this kind could ever take place again.
Digital Counter-Colonization
If digital colonialism is akin to the actual colonialism of the twentieth century, then what would digital counter-colonization look like? The tech journalist Karen Hao, the author of the tremendously insightful bestseller Empire of AI, likes to use this line as a kicker in speeches and op-eds: “when people rise, empires fall.” However pithily reassuring, the aphorism is misleading. Empires fall for many reasons besides mass insurrection, and what succeeds them often brings problems of its own.
Using history as a guide, the most natural successor to the empire would be the nation-state. In the 1970s and 1980s, Latin American countries were prominent advocates of a New International Economic Order to overcome the enduring legacies of colonialism. Using a language of national sovereignty, they also called for a New International Information Order to undo the dominance of the industrial North, which controlled telephone and telegraph wires, submarine cables, train tracks, electrical grids, the news services, and televised entertainment. Their solution was not a nebulous uprising of the people but a coordinated effort of governments to safeguard their autonomy in an interconnected world.
Digital counter-colonization would require many of the same things that actors in the Global South were asking for in the 1960s and 1970s: matching the promise of national sovereignty formally held out by international law with the capacity to exercise that sovereignty. Alongside digital empires, one could discuss a Digital Non-Aligned Movement and a Silicon Bandung to create space between the two poles of absolute refusal and absolute domination. Brazil’s legal pushback against the non-compliance of Musk’s X and recent push toward data localization and domestic data center construction under the banner of “digital sovereignty” represent efforts in this direction.
There are two reasons the critical literature scarcely calls for digital nationalism. The first is that the most potent efforts to contain the World Wide Web within national borders have come from authoritarian countries seeking to stifle the free exchange of ideas and any challenge to one-party rule. In the most prominent example, the so-called Great Firewall has inoculated Chinese society against external political interference while fostering replacements for Silicon Valley’s products in national champions like Baidu, WeChat, and Alibaba. With the exception of high-end AI chips, China has built the full stack of hardware, software, and apps to enhance its own national capacity against the incursions of would-be “digital empires”—but its own curtailment of political freedoms makes it a dim beacon for most progressive scholars and activists.
Another reason digital nationalism feels inadequate to many critics is that they are writing from a theoretical tradition in which the achievement of formal independence was a false dawn. For instance, the Peruvian sociologist Aníbal Quijano’s influential concept of the “coloniality of power” claims that in important ways colonialism never ended. The racialized hierarchies that were essential to the colonizing project persisted into the era of national independence. In this understanding, the very structure of Western rationality has always worked to discredit and extinguish indigenous forms of knowledge; digital technologies and platforms are merely the most recent carriers of the pathogen of modernity.
If, as Couldry and Mejias write, “the footprint of the new data colonialism will be overlaid upon the 500-year-old map of historical colonialism,” efforts at nationalizing data or exerting power over transnational tech companies can always be co-opted by a new cast of corrupt or power-hungry elites in a way that leaves basic power imbalances intact.
Rather than toggling between colonialism and nationalism, one way out is to swap political categories for economic ones. One scholar who does is Vili Lehdonvirta of the Oxford Internet Institute. Though his book is titled Cloud Empires, its most powerful insight is that there is something happening in our engagement with transnational digital platforms that most scholars fail to pick up on: an old-fashioned process of class formation.
As digital-gig workers find each other online, they start to develop tools for coping with often faceless and algorithmically driven platforms. Similarly, software engineers and project managers at companies like Google and Meta have expressed political opposition to the contracts with the Pentagon into which their employers have entered. These moments of class consciousness panicked the bosses and propelled many in the tech leadership class to the right.
Lehdonvirta analogizes such episodes to the process of state formation in the late Middle Ages. As European burghers learned to coordinate long-distance trade through private law and non-state networks of arbitration and credit, they began to flex their political influence by embarking on projects of self-governance in “free cities” like Bremen and Hamburg. What if, Lehdonvirta implies, a similar process of embourgeoisement were happening now through digital capitalism, with less empire creation than nation-state formation?
In this analogy, recognizing that we co-create value in our interaction with networked technologies—that we are digital creoles—becomes the first step toward making claims on the profits produced. Scholars and activists have proposed “data trusts” or “platform co-ops” as ways of reformatting relationships of ownership and decision-making. As Clarissa Redwine and J.S. Tan describe in a forthcoming book, the tech-workers’ movement eschews the mystifying language of serfdom and empire for the practical terminology of walkout and strike.
One interesting feature of the present moment is the spectacle of panicky Silicon Valley leaders speedrunning two centuries of political history to crack the riddle of securing social consent as they roll out what they promise themselves will be the radically disruptive technology of generative AI. These proposals range from drafting an AI constitution to innovations like social safety nets and a public wealth fund. Disingenuous or not, such efforts to renegotiate a social contract for the age of AI showcase something quintessentially modern: private firms are aware that their power finds its limit when they lose public legitimacy. As the techlash gathers steam and polls show public support for AI cratering, it is time to set aside the polemical language that has seemed so necessary recently to rouse the complacent masses. Against the pronouncement of one prominent tech critic, capitalism is not dead—nor do we live in literal empires. But this doesn’t make things any easier. As a recent film had it, it’s still just one battle after another.
Quinn Slobodian is professor of international history at Boston University. His most recent book is Muskism: A Guide for the Perplexed co-authored with Ben Tarnoff.
