DOI: 10.1093/9780197849644.003.0013 ISSN:

Governing Algorithmic Truth

Donghee Shin

Abstract

AI-driven misinformation detection has emerged as a critical site of algorithmic governance. Policymakers around the world are increasingly turning to regulatory mechanisms, such as the EU AI Act, algorithmic impact assessments, and ethical oversight frameworks, to confront the sociotechnical challenges of automated content verification. Fact-checking systems are better understood not merely as tools but as epistemic infrastructures that embed normative assumptions about evidence, credibility, and truth. This chapter evaluates governance mechanisms (algorithmic audits, contestability protocols, and independent oversight bodies) for their capacity to foster transparency, procedural fairness, and democratic accountability. Drawing on relevant scholarship and regulatory developments, it proposes a framework for reflexive, participatory, and democratically accountable oversight of AI fact-checking. It argues that governing algorithmic verification requires not only safeguards against technical harm but also structures that support plural truth practices and inclusive epistemic infrastructures.

More from our Archive