BiasPrivacyFairnesscs.CY
Should they? Mobile Biometrics and Technopolicy meet Queer Community Considerations
Anaelia Ovalle, Davi Liang, Alicia Boyd
October 6, 2023
Conference on Equity and Access in Algorithms, Mechanisms, and Optimization
5 citations
Key Takeaways
-
Biometric data from smartphones poses unique risks to marginalized communities: The paper highlights how gait recognition, gender recognition, and other biometric identification technologies built into smartphones can enable surveillance and discrimination against queer individuals, particularly given historical patterns of harassment and violence targeting this community.
-
Algorithmic bias in biometric systems can cause psychological harm: The research identifies "algorithmically-driven dysphoria" as a specific concern, where biometric systems that misclassify or misidentify individuals based on gender or other characteristics can cause direct emotional and psychological distress to queer users.
-
Current data protection practices are insufficient for vulnerable populations: The paper argues that existing approaches to biometric data collection and use don't adequately account for the sociotechnical realities and specific vulnerabilities of queer communities, suggesting that one-size-fits-all privacy policies fail to address community-specific harms.
-
Responsible AI requires centering marginalized perspectives in system design: The authors emphasize that developing ethical biometric technologies requires actively engaging with and considering the concerns of queer communities during the design and policy-making process, rather than treating these concerns as an afterthought.
-
New frameworks for biometric governance are needed: The paper calls for rethinking data protection practices and biometric collection standards to better safeguard human rights, particularly for communities with historical reasons to distrust surveillance technologies.