Dan M. Kotliar in “Data Orientalism: on the algorithmic construction of the non-Western other”, explores the concept of data colonialism and the algorithmic gaze along with the ways in which algorithms are programmed to see, depict, and understand the Other.
Scholars Couldry and Mejas compare the appropriation of territory, resources, and people for profit that characterized historic colonialism to data relations and the exploitation of human beings through the extraction of their data that defines data colonialism. According to Couldry and Mejas, what makes the collection of data colonial is its tendency for expansion, both geographic and social. “Data colonialism simultaneously seeks new territories to set its algorithmic eyes on, as well as new aspects of sociality that have previously been missing from its purview” (4). Personal data, in this sense, represent the new oil and currency in the digital media world. Data are continuously extracted offering opportunities for behavioral influence on users. Historic colonialism and data colonialism are based on the same principle of “cheap nature”. Historic colonizers used this principle to justify their actions since, in the place they appropriated, there was an abundance of natural resources and raw materials, whose appropriation was considered unproblematic from the colonizers’ point of view. The same principle is applied to personal data, that are abundantly produced by users and can be used. Moreover, the concept of social rationality gives a justification to this process of data extraction treating the users labor, which makes the data collection possible, as “just sharing”. Â
Kotliar emphasizes the idea that colonialism was not only about military force or processes of modernization, but it was also about knowledge creation - the way in which colonizers categorized indigenous into racial, ethnic, and cultural categories which sustained and justified colonial power. This cultural differentiation is what Edward Said described as Orientalism – the process about the creation and the distinction between Oriental and Westerns. “Said predominantly saw Orientalism as a discourse … that systematically devalues the Oriental Others, depicting them as backward, primitive, or weak” (6).
Kotliar bases his article on an empirical study on Extractive, an Israeli start-up company that provides user profiling algorithms to the East Asian market. It gains access to users’ Facebook account and makes the users transfer their credentials and their data into the companies’ hands. These profiles can be later used to create a market segmentation and approach costumers with personalized offers, contest, or services. Therefore, the Extractive algorithms transform anonymous users into accessible costumers. Moreover, Extractive works with Deep Pocket Inspection (DPI) which allows companies to gain access to users’ smartphones and their location, text messages, apps and many other features. Extractive argues that their work does not have cultural or linguistic barriers and have, thus, a universalizing perspective. Instead of categorizing people based on their racial, cultural, and ethnic identity Extractive divides them basing on personal preferences. They have a set of just 30 categories thar can describe people in the entire world. The categories it offers to its clients include names like “techie”, hipster”, “globetrotter”, “artsy” or “sporty”. Therefore, while they claim their algorithms can disregard language and universally categorize people, the categories they created are from a very culture-specific perspective – a Western point of view. Data colonialism, indeed, is deeply connected with capitalism and maximization of profit.
The implications of data colonized subject can be:
 -   Tethered of data judgment: algorithmic discrimination. For example, the face recognition technology perpetuates the racial bias. The algorithm decides, like in the HP case, whether a human face is recognizable or not.
- Â Â Â Dispossession through data: colonial appropriation. For example, Facebook can decide for a person of color to show the job positions that are more adequate based on who he/she. And, generally, are the job positions that are less prestigious compared to the white counterpart.
- Â Â Â Integrity of the self-evaluated by artificial intelligence. For example, the software Proctorio has been increasingly used to take exams to make sure students do not cheat taking their exams online during the pandemic. Most of the tools of artificial intelligence inside the software do not recognize black students, or anxiety symptoms like tics or spams.