Questioning Technological Hierarchies
How can facial recognition divide the use and analysis of marginalized communities?
Facial recognition technology is a great example of how bias can be built into systems. Research by Buolamwini and Gebru (2018) revealed that commercial facial recognition systems had significantly higher error rates when identifying dark-skinned individuals, especially women. These inaccuracies are not just technical flaws—they can have serious consequences. This issue acknowledges how marginalized groups often get the short end of the stick of technological advancements, mainly when those systems are used for policing and security, because they are simply not taken into mind during development.
Follow-up from question one: Who decides how to shape and create these boundaries?
Technology's design and ethical boundaries are often set by a small, homogeneous group—typically white, male, and from the Global North, like the United States. This lack of diversity can result in technologies failing to account for marginalized communities' needs and experiences. For example, voice recognition systems often struggle to understand non-standard accents or dialects (Tatman, 2017), excluding users outside the dominant linguistic norm, and specifically those in cities where English isn’t the dominant language. A more inclusive approach—such as participatory design involving the communities most affected—could lead to more ethical and equitable technologies. However, these perspectives are often sidelined in corporate tech environments, prioritizing profit and speed over inclusivity.
How do surveillance technologies unfairly target and criminalize marginalized communities?
Surveillance technologies are brought out unevenly, often targeting communities that are already over-policed and stigmatized. Predictive policing tools, for instance, use historical crime data—which already reflects racial bias—to predict future crime "hotspots," leading to a feedback loop of increased surveillance in poor and minority neighborhoods (Richardson, Schultz, & Crawford, 2019). Similarly, surveillance tools in immigration enforcement, such as biometric tracking and social media monitoring, disproportionately affect undocumented and racialized communities.
How do social media platforms' polices marginalize people of color while amplifying others?
Major platforms like Facebook, Instagram, and TikTok have also been criticized for disproportionately silencing marginalized voices. Activists—especially those advocating for racial justice, Indigenous rights, or Palestinian liberation—have reported having their content removed or accounts banned under vague or inconsistently applied guidelines (De Vynck & Zakrzewski, 2021). These algorithmic decisions are rarely transparent and often shaped by international politics or commercial interests, like more money for a company. As a result, dominant narratives are amplified, while critical voices are suppressed, reinforcing existing power dynamics.
References:
Buolamwini, J., & Gebru, T. (2018). Gender shades: Intersectional accuracy disparities in commercial gender classification. Proceedings of the 1st Conference on Fairness, Accountability and Transparency, 81, 77–91. https://proceedings.mlr.press/v81/buolamwini18a.html
De Vynck, G., & Zakrzewski, C. (2021, October 25). Facebook’s own data shows it suppresses Palestinian content. The Washington Post. https://www.washingtonpost.com/technology/2021/10/25/facebook-palestinian-suppression/
Richardson, R., Schultz, J. M., & Crawford, K. (2019). Dirty data, bad predictions: How civil rights violations impact police data, predictive policing systems, and justice. New York University Law Review, 94(1), 192–233. https://www.nyulawreview.org/issues/volume-94-number-1/dirty-data-bad-predictions-how-civil-rights-violations-impact-police-data-predictive-policing-systems-and-justice/ Tatman, R. (2017, March 21). Gender and dialect bias in YouTube’s automatic captions. Making Noise in the Machine. https://makingnoise.blog/2017/03/21/gender-and-dialect-bias-in-youtubes-automatic-captions/
