AccountabilityFairnessPrivacycs.HC
How Platform-User Power Relations Shape Algorithmic Accountability: A Case Study of Instant Loan Platforms and Financially Stressed Users in India
Divya Ramesh, Vaishnav Kameswaran, Ding Wang, Nithya Sambasivan
May 11, 2022
Conference on Fairness, Accountability and Transparency
61 citations
4 influential
Key Takeaways
-
Power imbalances undermine technical accountability measures: The study found that transparency mechanisms like audits and explainability—common approaches to AI accountability—may be ineffective when users are in vulnerable positions. In this case, financially-stressed Indian users accepted harmful loan terms despite risks including abuse, recurring debt, and privacy violations, suggesting that technical fixes alone cannot ensure accountability in unequal power dynamics.
-
Users internalize blame rather than holding platforms accountable: The research revealed that users "assumed responsibility for their negative experiences, thus releasing the high-powered loan platforms from accountability obligations." This psychological dynamic—where vulnerable users felt indebted to platforms for providing loans—prevented them from demanding accountability, even when experiencing documented harms.
-
Context matters: Global North solutions may not work in Global South settings: The authors emphasize that "the success of these mechanisms may be limited to Global North contexts," highlighting that responsible AI policies developed in wealthy countries may fail to protect vulnerable populations in different socio-political conditions. Policymakers cannot simply transplant accountability frameworks without considering local power structures.
-
Vulnerability creates exploitative dependencies: Users demonstrated persistent dependence on loan platforms despite known risks, including over-sharing sensitive data and paying high fees to unverified lenders. This suggests that when people are financially desperate, they become trapped in relationships where platforms can exploit them with minimal consequences.
-
Policymakers should move beyond purely technical approaches: The study cautions against relying solely on technical solutions to AI accountability, implying that regulations must address underlying power imbalances and socio-economic vulnerabilities to be truly effective.