ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

Strengthening democracy through AI governance innovations

Human Rights
Political Participation
Experimental Design
Christine Galvagna
University of California, Berkeley
Christine Galvagna
University of California, Berkeley

Abstract

This paper will explore how governance innovations can enable the public to exert meaningful democratic control over artificial intelligence (AI). Private companies developing AI and other digital technologies are responsible for an ever-growing list of consequential political decisions that would conventionally be made by elected representatives. For example, when designing and testing certain AI systems, companies must judge whether their system poses an “acceptable” level of risk to fundamental rights to comply with the EU’s Artificial Intelligence Act. They must do so individually or collectively in industry-led standardization bodies, with little or no guidance from public officials. Public officials lack the technical expertise, capacity, and/or political will to do so. Consequently, democratic participation by citizens is rendered less effective, as they have dwindling control over the increasingly privatized decisions that shape their lives. Governance innovations in the design and management of these technologies may introduce meaningful democratic control over otherwise private decisions. Research on co-creation and other models of participatory technology governance illustrate the possibilities of citizen or user control (see, e.g., Fan and Zhang, 2020; Robinson et al., 2021; Schneider et al., 2021; Zhang et al., 2020). This paper will investigate the feasibility and desirability of different forms of public participation in AI governance. The author will conduct interviews and roundtables with representatives of companies responsible under the AI Act for deciding what constitutes an acceptable level of risk to fundamental rights, as well as experts in human rights law and civil society representatives. Discussions will center on which forms of participation are most feasible in companies’ development and management processes, and which are most likely to empower participants to safeguard those rights. The paper will report these findings, and offer concrete recommendations for companies willing to implement pilot projects with citizen or user participation. In the absence of meaningful democratic control by elected representatives, these governance innovations may strengthen democracy by empowering the public to exert control over privatized decisions that impact their lives. Fan, J. and Zhang, A. (2020). Digital Juries: A Civics-Oriented Approach to Platform Governance. http://dx.doi.org/10.1145/3313831.3376293. Robinson, D., Simone, A, and Mazzonetto, M. (2021). RRI legacies: co-creation for responsible, equitable and fair innovation in Horizon Europe. https://doi.org/10.1080/23299460.2020.1842633. Schneider, N., et al. (2021). Modular Politics: Toward a Governance Layer for Online Communities. https://arxiv.org/pdf/2005.13701. Zhang, A., Hugh, G., and Bernstein, M. (2020). PolicyKit: Building Governance in Online Communities. http://dx.doi.org/10.1145/3379337.3415858.