BiasFairnessTransparencycs.HC
Filters of Identity: AR Beauty and the Algorithmic Politics of the Digital Body
Miriam Doh, Corinna Canali, Nuria Oliver
June 24, 2025
arXiv.org
1 citations
Key Takeaways
-
AR beauty filters are not neutral tools: The paper argues these filters function as "technologies of governance" that actively reinforce specific beauty standards rather than simply offering optional cosmetic enhancements, making their design choices a matter of responsibility and ethics.
-
Algorithmic bias perpetuates discrimination: The filters embed racialized, gendered, and ableist beauty standards through their underlying algorithms, meaning they systematically advantage certain appearances while disadvantaging others based on protected characteristics.
-
Hidden influence through naming and design: The paper identifies how naming conventions and platform governance choices obscure the filters' influence on users, making it difficult for people to recognize they're being exposed to narrow aesthetic norms presented as universal ideals.
-
Transparency is essential for accountability: The authors advocate for transparency-driven interventions as a key solution, suggesting users and regulators need clear information about how these filters work and what standards they encode to make informed choices.
-
Digital embodiment requires critical examination: The paper calls for rethinking how we approach algorithmic aesthetics more broadly, positioning this as part of a larger conversation about body politics and how technology shapes our relationship with our own and others' bodies.