Algorithmic colonialism

Algorithmic colonialism is Abeba Birhane's name for the pattern by which Western technology corporations impose AI systems on African populations following the structural logic of historical colonial appropriation: framing AI products as solutions for the developing world; extracting African population data as raw material; and institutionalising Western algorithmic assumptions into African social infrastructure — replicating, through corporate profit motives rather than political force, the three-stage movement of justification, extraction, and institutionalisation that characterised historical colonialism. The framing was introduced in a July 2019 essay in Real Life magazine and developed into a peer-reviewed paper in SCRIPTed: A Journal of Law, Technology and Society in August 2020, before being reprinted in the 2023 Oxford University Press volume Imagining AI: How the World Sees Intelligent Machines. Within the corpus, it is the canonical Africa-and-diaspora framing through which the AI-accountability and decolonial-tech movements identify the structural harm of AI deployment as an extension of extractive colonial power, and it is distinct from the data colonialism framing used by Latin American civil society: where data colonialism addresses the general asymmetry of data accumulation between the Global North and Global South, algorithmic colonialism names the AI-and-corporate specificity of how that asymmetry plays out on the African continent and in the African diaspora.

Origin

Birhane developed the framing at the intersection of her academic work in cognitive science at University College Dublin and her own biographical position as an Ethiopian-Irish researcher — someone whose life sits at the Africa-diaspora nexus the framing addresses. The July 2019 Real Life essay was the public debut, framing AI deployment in Africa as following colonial extraction patterns through a three-act structural logic: first, the justification act, in which Western technology corporations frame their products as "solutions for the developing world", treating complex, historically and culturally rooted African social problems as amenable to algorithmic fix; second, the extraction act, in which large-scale indiscriminate data collection — describing African populations as a "data-rich continent" in language Birhane flags as echoing colonial resource-mining vocabulary — takes populations as "passive objects for manipulation" rather than as agents; third, the institutionalisation act, in which Western algorithmic systems designed around Western values, cultural assumptions, and corporate interests are embedded into African social infrastructure without local adaptation, leaving the continent dependent on the corporations that built them.

The peer-reviewed version, published in SCRIPTed vol. 17, issue 2 (August 2020, pp. 389–409), set out the full argument with the working formulation that has since become the framing's canonical short form: "colonialism in the age of Artificial Intelligence takes the form of 'state-of-the-art algorithms' and 'AI driven solutions' to social problems" — control operating through "invisible and nuanced mechanisms" rather than brute physical force, driven by corporate agendas rather than government authority. Birhane's evidence included Nigeria's approximate ninety-percent software import dependency from Western corporations, the failure of AI tools built on Western training data to function accurately in African contexts, and the structural harm of FinTech and microfinance "solutions" that promised poverty alleviation while routing value extraction to foreign investors. The framing argued not simply that AI tools perform poorly in African contexts but that this underperformance is not accidental — it is the predicted outcome of a design process that never incorporated African populations as subjects with standing, only as raw-material sources.

Propagation

The framing's propagation ran along two tracks from 2020.

Academic channels

The SCRIPTed paper became a frequently-cited reference in the African-AI scholarship field. In 2023, it was reprinted as a chapter in Imagining AI: How the World Sees Intelligent Machines — an Oxford University Press anthology edited by Stephen Cave and Kanta Dihal that positioned the algorithmic colonialism framing alongside Global South AI perspectives from Latin America, South Asia, and East Asia in a comparative scholarly frame. Between 2021 and May 2023, Birhane served as a Mozilla Senior Fellow in Trustworthy AI, and subsequently as a consultant to Mozilla Foundation and Adjunct Professor at Trinity College Dublin, extending the framing's reach through Mozilla's civil-society AI-governance networks and through academic channels in computational social science and AI ethics.

Pan-African journalism and civil-society media

The Elephant, a Nairobi-based pan-African analysis, opinion, and investigation platform, republished the paper on 21 August 2020, distributing the framing across the English-reading African civil-society and academic field beyond the specialist law-technology journal in which the peer-reviewed version appeared. ETC Group — the international civil-society technology-assessment organisation that has tracked corporate concentration in food, agriculture, and life sciences for decades — produced a dedicated podcast episode ("Algorithmic Colonisation with Abeba Birhane," episode #5) in which Birhane set out the framing for a civil-society audience. The Radical AI Podcast (episode 11: "Robot Rights? Exploring Algorithmic Colonization with Abeba Birhane") extended the framing's reach into US-and-diaspora AI-ethics organising, reaching the academic-adjacent AI-accountability community that has engaged most deeply with the adjacent coded gaze framing from that direction.

Adjacent framings in African civil society

The algorithmic colonialism framing does not operate in isolation in the African AI-accountability field; it sits alongside two parallel civil-society framings from in-corpus organisations that operate in the same intellectual register without using the same term.

Pollicy, the Kampala-based afro-feminist civic-tech collective founded by Neema Iyer, produced in 2021 the parallel framework Automated Imperialism, Expansionist Dreams: Exploring Digital Extractivism in Africa — identifying nine forms of digital extractivism operating in Africa (data extractivism originating with Western corporations and increasingly adopted by local ones among them) and arguing that "the structures and systems that support algorithmic colonialism" can be countered through strong data-governance and digital-rights policy. The Pollicy formulation is a direct African civil-society parallel: where Birhane's academic framing names the structural logic of corporate AI as colonialism, Iyer's practitioner framing names the specific extraction mechanisms that instantiate it and the governance responses that can counter it.

Paradigm Initiative, the Lagos-headquartered pan-African digital-rights non-profit, deploys the adjacent "algorithmic apartheid" framing in its advocacy — warning that algorithms trained on unrepresentative data risk "enveloping and trapping" African communities in AI apartheid systems, and calling for algorithmic auditing frameworks as the governance response. The framing is substantively kin to Birhane's but specifically names South African apartheid rather than classical colonialism as its historical referent, and routes through Paradigm Initiative's Digital Rights and Inclusion Forum (DRIF) as the primary pan-African civil-society convening surface.

The Distributed AI Research Institute (DAIR), founded by Timnit Gebru in December 2021 with an explicit mandate to address AI harms to Africa and the African diaspora through community-rooted research, is the diaspora-side institutionalisation of the intellectual tradition Birhane's framing names. DAIR's community-rooted research model — explicitly designed to counter the extractive relationship in which African populations bear AI's harms while AI's benefits flow to the corporations and research institutions that built the systems — is the organisational form through which the diaspora half of the algorithmic colonialism framing operates as a standing research programme.

Why it differs from data colonialism

Three features distinguish the algorithmic colonialism framing from the adjacent data colonialism framing that Latin American civil society (Derechos Digitales, Coding Rights) deploys:

First, the algorithmic colonialism framing is AI-and-corporate-specific where data colonialism is structural-systemic. Where data colonialism names the general asymmetry of accumulation between jurisdictions — any data extraction by any corporate or state actor — algorithmic colonialism names the specific mechanism through which that asymmetry operates in the AI age: the three-stage justification-extraction-institutionalisation cycle of corporate AI deployment, with its particular emphasis on the inappropriateness of AI "solutions" designed for one demographic and imposed on another.

Second, the framing is Africa-specific rather than Global-South-general. Birhane's argument draws on African case material — Nigeria's software import dependency, FinTech deployment patterns, the failure of face-recognition systems built on majority-white training data when deployed in African contexts — rather than on a cross-regional structural-economic claim. This specificity makes the framing more granular but also more bounded: it is a claim about a particular continent and its particular relationship to Western AI corporations, not a universal claim about data extraction.

Third, the framing explicitly addresses the African diaspora as a co-subject of the harms it names — not only Africans on the continent harmed by AI systems deployed there, but also Africans in the diaspora harmed by AI systems designed in the same Global North jurisdictions and deployed in contexts (European policing, US hiring and credit systems) where the diaspora is also present as a population. This double address — both the continent and the diaspora — is what the adjacent coded gaze framing (originating with diaspora scholar Joy Buolamwini and anchored in the Algorithmic Justice League) shares: both framings speak from and to a Black-and-African-diaspora position within the AI-accountability movement, from institutional homes in Ireland and the United States respectively, and both address AI harms that affect the same populations on both sides of the Atlantic simultaneously.