A review of Mejias and Couldry, “Data Grab: The New Colonialism of Big Tech and How to Fight Back” (University of Chicago Press, 2024).
Last year, I carried Ulises A. Mejias and Nick Couldry’s new book “Data Grab” with me to two formerly colonizing and two formerly colonized countries. In Amsterdam’s imposing Rijksmuseum, my Dutch-Indian friend said, “You’ve seen the beauty; now comes the horror,” and he led me to the painting of a Dutch sugar factory in Java, with a slave bell for the enslaved workers. He showed me beautiful seascapes of ships and said that one of the museum’s exhibits marked out several as slave ships. Without the annotations and commentary, I might have missed the violence hidden in the art. For the tech industry, Couldry and Mejias do the same kind of valuable work my friend did for me in the museum, by highlighting the colonial, extractive violence that underlies and defines our digital lives, both historically and in the present. They started this project in a co-authored article and in their previous co-authored book, “Costs of Connection.” Their latest book, titled “Data Grab” to invoke colonialism’s land and resource grabs, continues their compelling work. “Data Grab” presents a more accessible, reorganized exposition of the authors’ views.
Mejias and Couldry’s account of the capitalism-driven “social quantification” sector (as they call it) fills gaps in Shoshana Zuboff’s compelling and popular “The Age of Surveillance Capitalism.” Zuboff did the valuable work of making mainstream the idea that we, the datafied subjects, stand in an asymmetric relationship with “surveillance capitalist” companies. Although powerful and influential, Zuboff’s account of the Big Tech companies’ business models does not engage with how far their harmful characteristics come from capitalism, their disparate impact on different groups of people, or their reliance on law for their power. Julie Cohen has offered a magisterial account, framed as “informational capitalism” (from Manuel Castells’s “The Rise of the Network Society”), covering this ground in a book more suitable for academic readers. However, Zuboff’s book captured the public imagination, and her discussion of colonialism is confined to Britain’s relationship with the United States, eliding the slavery and worldwide extraction and exploitation integral to European colonialism.
“Data Grab” fills the gap in Zuboff’s account of colonialism by discussing the connection between capitalism and colonialism and engaging with colonialism’s complex, transnational nature. Mejias and Couldry focus on the informational dimension of what they call the “new colonialism,” drawing valuable connections between historic colonialism and “data colonialism.” I discuss some important contributions from their work below, including what “data grab” means and the implications of this new colonialism layered over “historic colonialism.” I also discuss how law enables companies to “explore, expand, exploit, and exterminate” (as Mejias and Couldry describe it) since legal ordering is a part of the institutional capture that enables data colonialism. “Data Grab” is a great resource for professors, researchers, advocates, and others interested in the political economy of the major technology platforms.
Grabbing Data
In their prior work, Mejias and Couldry argued that technology companies’ appropriation of “many specific aspects of human life” is best understood as “data colonialism.” Mejias and Couldry don’t use colonialism as a metaphor. They present data colonialism as the “new era of coloniality” and argue that it shares the core missions of earlier colonial endeavors, namely, extraction or dispossession, which they suggest takes place in four stages: exploration, exploitation, expansion, and extermination. In “Data Grab,” the authors clarify that data colonialism is not a “literal comparison” to (historic) colonialism but a “new colonialism,” which they offer as a framework to understand our digital lives and the relationships to corporations that make these lives possible. In doing so, they argue that “historic” and “new” colonialism are parallel and linked processes.
“Data Grab” invokes land grabs, and the “extraction” metaphor for data, by comparing data to land and natural resources. This can be misleading because it suggests that data has a “raw” form. The authors are too knowledgeable to believe this: They identify the “imposition of one reading of how the world and its knowledges should be organized” as one of the core problems of data colonialism. This understanding might have molded Mejias and Couldry’s use of the extraction metaphor to account for how data is constructed. The authors acknowledge that “the classification of people into disadvantageous categories” is among data colonialism’s harms, but they attribute it to “AI and its algorithms” rather than to datasets. Datasets remake people’s social identities by classifying them in particular ways, whether by gender, race, or something else. Arguably, the violence of recording data is less like stealing a community’s land or oil, and more like rebuilding towns without piers for a seafaring community, or restrooms that accommodate only people of one gender or people whose mobility is unimpaired.
Highlighting classification as a feature of data colonialism would reinforce Mejias and Couldry’s discussion of the ideological foundations of old and new colonialisms. Colonialist ideologies defined and justified who exploits and who is exploited through race science and racialized classification. The authors acknowledge the history of these exploitative relationships and the ideologies driving them, and might have discussed how this also affected how data is classified. Mejias and Couldry do allude to this briefly, by discussing “data violence” and how discriminatory datasets produce algorithmic discrimination, but a more detailed explication of how this fits the colonial metaphor would have been helpful.
That said, Mejias and Couldry’s narration of how technology companies accumulate data, creating capitalism-driven systems intended to maximize profits by commodifying social engagement, is insightful. Their examples are compelling, especially when they offer illustrations of how platforms choose monetization over desisting from harm, in hosting highly profitable nonconsensual sexual and violent media instead of enabling the people harmed by this content to take it down easily.
Data’s Colonialisms
While the informational dimension of data colonialism is important, it is worth noting that data colonialism both reinforces older harm and inflicts new injuries. The authors acknowledge this, writing of “this strange combination—the continuation of colonialism’s inherited inequalities and colonialism’s acquisition of new tools that potentially affects human life in radically new ways.” Colonialism’s inherited inequalities order where people are in data’s value chain. In addition to data colonialism’s informational harms, people such as exploited data workers and physical laborers, who are often erased from discussions of the data economy, experience material harms. Informational capitalism relies on exploited labor, including children risking their lives mining cobalt in the Congo because capitalism has left their families with no other options. The new colonialism builds on the old one, and the wounds of historic colonialism have not healed.
The authors highlight “efficiency of extraction,” a key feature of colonialism, which used productive methods “at the cutting edge of nineteenth-century accounting, financing and the ‘rational’ management of human bodies to produce economic output on the largest possible scale and at the lowest possible cost.” They draw on this feature—the system and ideology that supported colonial extraction—to support their framing of Big Tech corporations’ presence in our digital lives as colonialism. Mejias and Couldry also discuss the colonial origins and organization of platformized work and how BigTech platforms excel at controlling and exploiting vulnerable workers across the world, especially workers from groups previously victimized by earlier rounds of colonialism. This is a part of the material effects of data colonialism. Platforms and data are used to control and exploit workers, as scholars such as Ifeoma Ajunwa, Veena Dubal and Karen Levy have shown. This includes workers who clean and label datasets.
But there is more to this story. Poorly paid Kenyan workers labeled data to make ChatGPT less violent, sexist, and racist in dominant languages. The large language model is unlikely to be proficient in their local languages. Data colonialism’s extraction covers more than data and exploited labor. Like historic colonialism’s inequalities, it also devalues and extracts labor in the form of service and care work associated with women, including the erased familial systems of care that sustain outsourced workers Through her interviews with call center workers and Indian computer engineers, Kalindi Vora points out that transnational capitalism and the colonialist division of labor form a system that denies southern workers (typically from populations disempowered by colonialism, for example in India or Kenya, but also including migrant labor or descendants of enslaved people in the United States or Europe) the stability, creativity, and capacity to engage in work they see as meaningful.
“Data Grab” takes the important step of highlighting how historically inflicted inequalities have affected where the companies owning globally dominant platforms are geographically located, and how post-colonial geopolitical relations allow “Global North” companies to extract from the “Global South.” Historic and new colonialism run “side by side” in their account. I would modify this slightly to argue that the new colonialism builds on the old one, but I suspect that Mejias and Couldry would agree. Perhaps the reason they discuss the two in parallel is to recognize that the wounds of historic colonialism have not healed and that data colonialism both reinforces them and inflicts new injuries.
The Law’s Heroic Role
Both Mejias and Couldry are scholars of communication, and perhaps that explains their optimistic view of the law’s role in data colonialism. Most references to the law in the book are to data protection law or how lawsuits can enable workers to assert their rights. Legal scholars take a more complex view of the law’s engagement with Big Tech companies, recognizing that although law can restrict the companies’ exploitative and extractive behavior, it can also enable it. Julie Cohen offers the most detailed and complex articulation of this view in her book, “Between Truth and Power,” in which she points out that institutional transformation of the law is a key part of the shift to the neoliberalism that drives the information economy.
The colonial land grab that Mejias and Couldry evoke was enabled by law, as was slavery for as long as law treated certain groups of people as property. The tech companies lobby legal institutions to create law that enables their exploitation, such as California’s Proposition 22, which allows companies such as Lyft and Uber to classify their workers as independent contractors and thus avoid the application of employee protection laws. The lack of adequate data protection laws permits the companies to gather data with impunity; and private law doctrines, such as the law of contracts and trade secrets, then permit them to hold and exercise exclusive control over this data.
The major tech companies influence law and legal discourse at a transnational level as well— lobbying international institutions such as the World Trade Organization, as well as lawmakers and adjudicatory institutions around the world—to create legal frameworks and interpretations of law that enable their business models. This is why I have suggested that corporate imperialism is a helpful framework for understanding their impact on the world. Data colonialism is made possible by a transnational legal order. Mejias and Couldry have excellent suggestions about local movements and activism to influence the law. Local movements will struggle, however, unless they are able to join hands across borders to ensure that companies’ harmful products (such as facial recognition technology) and harmful practices (such as classifying workers as independent contractors) do not proliferate internationally.
Conclusion
Mejias and Couldry offer a very readable and powerful account of Big Tech companies’ extractive practices and how they build on colonial extraction. They end on a hopeful note, with ideas for bottom-up resistance, including the example of Hiku Media, a data-driven project through which Maori people took the lead in preserving their own language, knowledge, and tradition, recording and defining them on their own terms instead of allowing the companies to define their culture by recording a distorted view of it.
I hesitate to ask more of this prolific duo, who have already offered up insightful frameworks to help us understand our datafied society and contest its oppressive features. It might have been like asking my friend at the museum why he didn’t discuss the women on the slave ships, after all I’d learned from his narrations of what I couldn’t see. Mejias and Couldry show how the legacy of colonialism means that harms are unevenly distributed even though everyone is at risk eventually. That is valuable. Perhaps their third book will spotlight the other harms data colonialism inflicts.
Author’s note: I am grateful to Julie Cohen, Nishant Shah, Jonathan Cedarbaum, and Yale Information Society Project ISP fellows Anat Leshnick, Elle Rothermich, Maria Angel Arango, Isaac May, and Michael Mc Govern for their feedback. All errors are mine.
– Chinmayi Arun is the executive director of the Information Society Project and lecturer in law at Yale Law School. Arun recently authored “The Silicon Valley Effect” (forthcoming in the Stanford Journal of International Law, March 2025). Published courtesy of Lawfare.
