Artificial Intelligence’s condition of possibility is coloniality. It’s development within the colonial matrix of power ensures it becomes a tool for dispossession and racialized and gendered violence.

Rachel Adams, “Can Artificial Intelligence Be Decolonized?” Interdisciplinary Science Reviews 46, no. 1-2, p. 176-178

Yet none of these texts, which constitute some of the leading work in the field, take into account the complex genealogy of intelligence: whose conception of intelligence is modelled within technology or how the idea has been put to work in dividing people between the desired and the undesirable. Nor the history of the human body as machine and commodity borne from slavery and colonialism, such that Achille Mbembe names blackness as the prototype for the assemblage of the human–object of modernity (2017, 2019). Nor the ways in which the knowledges upon which AI is built – statistical enumeration of people and land – were advanced by imperial powers to control and contain colonial populations (Appadurai 1993; Breckenridge 2014; Kalpagam 2000, 2014; Said 1978), and led to the development of cybernetics and eugenics, as well as the idea that, through feedback mechanisms, both human and machine can be corrected and improved. Indeed, for Mbembe we are encountering a third shift in the arrangement of race and blackness in global society: the first being slavery and colonization; the second being the development of writing and text, culminating in the formal processes of decolonization; and the third being the advent, proliferation, and ubiquity of digital technologies which represent the latest phase of high-modernity (2017). Similarly, for Aníbal Quijano, we are reaching a watershed moment in the global coloniality of power, with ‘the manipulation and control of technological resources of communication and of transportation in order to impose the technocratization/instrumentalization of Coloniality/modernity’ (2017, 364), together with ‘the mercantilization of subjectivity and life experiences of individuals’ (2017, 365). AI, too, is in the midst of reconciling three forms of discontent, 1 whose relation to the forms of coloniality and the historical construction of race at work in the world today has yet to be fully understood. First, manifest racial and gender bias within AI technologies (Benjamin 2019; Buolamwini and Gebru 2018; Keyes 2018; Noble 2018). In June 2020, E. Tendayi Achiume, United Nations Special Rapporteur on contemporary forms of racism, racial discrimination, xenophobia, and related intolerance, issued a report to the United Nations Human Rights Council that found ‘emerging digital technologies exacerbate and compound existing inequalities, many of which exist along racial, ethnic and national origin grounds’ (paragraph 4). The second discontent, echoing Quijano’s description of the mercantilization of life, regards the way AI and digital technologies are globally commoditizing human experience (Zuboff 2018). Within this data paradigm, the human is substituted as an assemblage of their data points which are, in turn, taken as a sign of the real (Baudrillard 1994). The third discontent is geopolitical, pertaining to the emergence of an arms race characterized by the ambitions of transatlantic nation states to be ‘global leaders’ (Vladimir Putin, cited in Vincent 2017) in AI innovation (Garvey 2019). 2 This, together with the second discontent, is central to the ideas of ‘data colonialism’ and ‘digital colonialism’ naming, within current AI discourse, the brazen flurry to extract and exploit personal data and data systems (Birhane 2019; Couldry and Mejias 2019; Crawford et al. 2019; Ricaurte 2019; Mhlambi 2020). It relates as well to Ian Hogarth’s critique of ‘AI nationalism’ where new dependencies are being tacitly enforced between low-tech and advanced-tech states and are set to follow the historical divide of Global North/South (2018). Perhaps more critically, the drive to dominate in the production and use of AI is revealing of the project’s hegemonic impulses, and the neo-Darwinian linearity of the evolution of science which will ‘leave behind’ those who do not conform to catching up. In light of these concerns, Mbembe’s provocation to critique the apparatuses of race and blackness in digital technologies, and Quijano’s warning against the hegemony of technocratization which encompasses, at once, the consolidation of modes of commodification of the human-object, becomes ever more urgent.