Ilmu Sosial editorial
Open AccessOA2024
Building, Shifting, & Employing Power: A Taxonomy of Responses From Below to Algorithmic Harm
Researchers synthesize historical cases into a taxonomy of responses 'from below' to algorithmic harm, connecting them to theories of power. The taxonomy highlights how people employ, shift, and build power, opening an opportunity space for the FAccT community to engage with and support such action.
Alicia DeVrio; M. Eslami; Kenneth Holsteinยท Conference on Fairness, Accountability and Transparencyยท 2024ยท DOI 10.1145/3630106.3658958
The core problem
Research on algorithmic fairness and transparency has largely focused on ensuring that algorithmic systems adhere to notions of fairness and transparency. However, a growing body of work emphasizes that mitigating algorithmic harms requires explicitly accounting for power structures. Those who wield power over algorithmic systems often fail to adequately address harms and rarely consult those directly affected. Consequently, individuals and communities respond to algorithmic harms in diverse ways, yet there is a lack of overarching understanding of these responses. This paper addresses that gap by synthesizing documented historical cases into a taxonomy of responses 'from below' to algorithmic harm. The taxonomy connects different types of responses to existing theorizations of power from anthropology, human-computer interaction, and communication, centering on how people employ, shift, and build power. The authors aim to highlight an opportunity space for the FAccT community to engage with and support such action from below.
Innovation
The authors conducted a synthesis of documented historical cases of responses to algorithmic harm. They drew from a wide range of sources to identify instances where individuals or communities reacted to perceived harms from algorithmic systems. Through iterative analysis, they developed a taxonomy that categorizes these responses. The taxonomy is grounded in existing theorizations of power from anthropology, human-computer interaction, and communication. The authors focused on how responses involve employing, shifting, or building power. The methodology likely involved coding cases and identifying patterns, though specific details of the synthesis process are not provided in the abstract. The resulting taxonomy aims to provide a structured understanding of the diverse ways people resist algorithmic harms.
Introduction
Research on algorithmic fairness and transparency has largely focused on ensuring that algorithmic systems adhere to notions of fairness and transparency. However, a growing body of work emphasizes that mitigating algorithmic harms requires explicitly accounting for power structures. Those who wield power over algorithmic systems often fail to adequately address harms and rarely consult those directly affected. Consequently, individuals and communities respond to algorithmic harms in diverse ways, yet there is a lack of overarching understanding of these responses. This paper addresses that gap by synthesizing documented historical cases into a taxonomy of responses 'from below' to algorithmic harm. The taxonomy connects different types of responses to existing theorizations of power from anthropology, human-computer interaction, and communication, centering on how people employ, shift, and build power. The authors aim to highlight an opportunity space for the FAccT community to engage with and support such action from below.
Why it matters
The taxonomy reveals that responses from below are not merely reactive but are active expressions of power. By employing, shifting, and building power, individuals and communities challenge the status quo of algorithmic governance. The authors argue that the FAccT community has an opportunity to engage with and support such action from below. This could involve amplifying marginalized voices, designing tools that facilitate collective action, or advocating for policies that enable power shifting. The paper contributes a broader understanding of how people resist algorithmic harms, moving beyond individualistic notions of fairness to collective and political action. It calls for researchers to consider power dynamics explicitly and to collaborate with affected communities in mitigating harms.
Who should read this
Pembaca Research Digest dan praktisi rumpun ini.
Opening member contentโฆ