The AI industry perpetuates historical patterns of othering and violence by reinforcing colonial, capitalist, and data exploitation systems that heavily surveil and control marginalized communities.
Tacheva & Ramasubramanian 23 [Jasmina Tacheva is an Assistant Professor in the School of Information Studies. Her research examines the mechanisms through which user-generated content shapes human behavior, with a focus on social justice, digital volunteerism, and sustainability. rividya “Srivi” Ramasubramanian holds a Ph.D. in mass communication from Pennsylvania State University. She is a leading communication scholar whose work addresses contemporary global issues relating to media, diversity and social justice. Ramasubramanian is widely recognized for her pioneering work on race and media, media literacy initiatives, implicit bias reduction and scholar-activism. Her areas of expertise include critical media effects, data justice, digital cultures, intersectionality, applied communication and mixed methods; Advance; “Challenging AI Empire: Toward a Decolonial and Queer Framework of Data Resurgence”; 03/02/23; https://advance.sagepub.com/doi/full/10.31124/advance.22012724.v1]
Theorizing the Faces, Roots, and Othering Violence of Al Empire The centuries-old tendency of Western epistemology to create a division between a Subject and a "Other" is at the heart of the violence inflicted by the unintended consequences of new media and technology, including Al systems (Arendt, 1958). Non-Western, non-white, non-male, poor, queer, transgender, dis/abled, non-Christian, non-capitalist, un-educated, subaltern, un-housed, un-free: people outside the dominant group are often perceived to be an obstacle in tech-progressive utopias and discarded as "surplus humanity," an afterthought, an impediment (Said, 1979; Muñoz, 1999). The Other is described as un-technological, or as only using technology for nefarious purposes, and therefore uncivilized, even though these communities are in truth the most heavily technologized for the purpose of their surveillance and control (Abdurahman and Ghoshal, 2021; Mavhunga, 2014). Dismantling the idea of the Other is no simple task, and it is therefore critical to keep in mind the caveats issued by transnational feminism reflected in the works of scholar-activists such as Chandra Mohanty. Combating the processes of othering marginalized communities cannot happen in the form of superficial empty signifiers such as "universal sisterhood," in the case of non-Western and Western women, because it is undeniable that there are significant differences in the conditions of these groups (Mohanty, 2005). These dehumanizing differences are brought about in no small part by the joint projects of coloniality, heteropatriarchy, and global racial capitalism the persistent sociopolitical, economic, and cultural order that engenders AI developments such as The Line and which is in turn shaped by them. The Line in Saudi Arabia, as well as the facial identification anti-protest crackdowns in Hong Kong, Iran, and Russia, along with NATO's "smart" weapons: The neo-imperialist race for power fueled by Al is so reminiscent of the imperial games of the past that leading political scientists have started talking about "Empire Resurgent," even though we live in ostensibly "postcolonial" times.? 2.1 Interlocking Matrices of the Al Empire: Data Colonialism, Data Commodification, Biometric Profiling, and Technological Apartheid As the world eagerly awaits the advent of a fully datafied city like The Line, where every action will be captured and translated into data in real time, the Uyghur people in the Chinese Xinjiang province one of China's largest ethnic and religious minorities— have had to live in a state of similar constant data capturing for years. Smart technologies enable the total integration of multiple data streams and sensorial information to surveil the online communication of journalists and dissidents critical of the monarchy in the form of military-grade "mercenary spyware" (House, 2022). The process by which governments and corporations privatize, weaponize, and claim ownership of the data produced by their users and citizens has come to be known as data colonialism (Coleman, 2018; Couldry and Mejias, 2019). In applying a colonial lens to the contemporary phenomena of datafication and data extraction, we should be careful not to erase the ongoing trauma and injustice brought about by the structures of modernity/coloniality, or the co-constitutive forces of colonialism and capitalism (Mignolo, 2007). The narrative of colonialism and modernity has come to shape many aspects of today's world through the material and epistemic violence issued by the West against its formal and informal colonies (Appiah, 2023). Much of today's state and corporate power relies on churning through large swaths of continuously captured data for decision-making and innovation. A big part 3 of these data streams are generated in former colonial spaces or made sense of through the labor of non-Western subjects. Take, for example, the recent revelations about the much-acclaimed OpenAI ChatGPT public user interface, whose remarkable properties are in part due to the $2 per hour labor of Kenyan content moderators.3 Once the data grab has occurred under the auspices of data colonialism, the context of data the human stories and circumstances keeping it tethered to the social world- is irretrievably lost, and data is turned into a commodity to be bought and sold in the Big Data marketplace of AI Empire. This process of data commodification stands in direct contrast to what makes us human: our sociality. Sociality implies building and maintaining relationships, not just within but also across identities, generations, and species. The neatly packaged data bundles of unified depersonalized aggregate data in which any trace of intersubjectivity is erased is just another product in the hands of the data brokers responsible for what Lamdan calls "the Costcoization of data" (Lamdan, 2022). This is the idea that data, even in its most diverse forms, is so ubiquitous and easy to find in one-stop data shops that acquiring it is akin to going to a big-box store to purchase everything you need in one place. Following this convenient data acquisition process is the analysis step, which enables Al systems to make predictions and categorize objects or humans for the purpose of determining their role and place in the world. This includes racial profiling and even the inference of human sexuality. In this context, the Uyghur experience in China is more properly understood as biometric mass surveillance, an extreme yet growing type of biometric profiling. This describes the use of systems or technologies that collect, analyze, and/or generate data (including biometric data) on large numbers of people instead of limiting surveillance to individuals about which there is 'reasonable suspicion' of wrongdoing (De Hert and Bouchagiar, 2022). Various streams of personal data such as skin color, height, face scans, and voiceprints from street surveillance cameras, and recently also mandatory DNA samples, are all fed into government integrated joint operations platforms where they can be linked to people's national identification numbers to yield a unique and comprehensive personal profile (Xu et al., 2022). Similar human rights concerns have been raised with respect to the introduction of the Unique Identification Authority of India's (UIDAI) Aadhaar, the world's largest biometric surveillance system, which has been critiqued as embodying Hindutva and neoliberal ideology aimed at surveilling and controlling Muslim, caste-oppressed, Indigenous minorities such as Dalits and Adivasis (Prabhakar, 2020). The adoption of cutting-edge Al-powered technology for state management and control by non-Western countries is often cloaked in the language of tech-progress. A case in point is the United Nation's enthusiastic welcome of Zimbabwe and Kenya's decision to introduce biometric-based identification, described as "leapfrogging] outdated tools and progressing] to modern innovations" (Toesland, 2021).