This paper argues that the racialised extraction of value is central to education technology that uses forms of artificial intelligence. Rather than suggest that racialisation is a problem to be ameliorated in EdTech, this paper contends that racialisation is essential to both the operation of AI supported EdTech, and its capacity to garner market share. To make this argument the paper has two parts. The first part of the paper outlines the centrality of racialisation to the operation of AI supported Edtech, focusing on two areas that underpin systems that use facial recognition technologies. The first area examines the links between race and training sets, including issues of exclusion and misrecognition of people of colour (Crawford & Paglen, 2021). This is now a common focus where addressing bias is seen as a key remedy for racial bias. Conversely, this paper draws on work that highlights that including people of colour in training sets can create more accurate systems, but not less pernicious ones as inclusion can be deleterious for historically marginalised and surveilled populations (Benjamin, 2019). The second area is algorithmic. While most focus on the links between race and technology are on data, there is an important but underexamined dimension in the historical racialisation of machine learning methods and algorithms. This includes machine learning methods used in common AI technologies in EdTech such as facial recognition. For example, facial recognition technologies use the Mahalanobis similarity measure which has racial origins in colonial rule in India (Taylor, Gulson, & McDuie‐Ra, 2021). The second part of the paper focuses on the notion of racialised extraction of value, drawing on critical theories of race and technology, including those related to racial capitalism (McMillan Cottom, 2020). This notion of racial capitalism provides insights in this paper to the way education technology derives both social and economic value through racialised data and algorithmic practices (e.g., Henne, Shelby, & Harb, 2021). In education technology this can include the production of market value, simultaneous with the production of allocative and representational harms, such as racial profiling while using education platforms (Nichols & Garcia, 2022). This paper concludes by contending that the racialised extraction of value is both necessary for including people of colour in Edtech (e.g., being able to use the products in the markets of the majority, non-White world), and yet also reinforces the historical and pernicious surveillance of people of colour in education.

Benjamin, R. (2019). Race after technology: Abolitionist tools for the new Jim Code. Cambridge: Polity. Crawford, K., & Paglen, T. (2021). Excavating AI: the politics of images in machine learning training sets. AI & SOCIETY, 36(4), 1105-1116. doi:10.1007/s00146-021-01162-8 Henne, K., Shelby, R., & Harb, J. (2021). The Datafication of #MeToo: Whiteness, Racial Capitalism, and Anti-Violence Technologies. Big Data & Society, 8(2), 20539517211055898. doi:10.1177/20539517211055898 McMillan Cottom, T. (2020). Where Platform Capitalism and Racial Capitalism Meet: The Sociology of Race and Racism in the Digital Society. Sociology of Race and Ethnicity, 6(4), 441-449. doi:10.1177/2332649220949473 Nichols, T. P., & Garcia, A. (2022). Platform Studies in Education. Harvard Educational Review, 92(2), 209-230. doi:10.17763/1943-5045-92.2.209 Taylor, S., M., Gulson, K. N., & McDuie‐Ra, D. (2021). Artificial Intelligence from Colonial India: Race, Statistics, and Facial Recognition in the Global South. Science, Technology & Human Values, 1-27. doi:10.1177/01622439211060839