Abstract
GenAI is rapidly becoming embedded in research, teaching, and public engagement, yet its role in shaping scholarly discourse remains contested. Of particular concern are GenAI systems’ content moderation mechanisms, which, while designed to prevent harm, often intervene in ways that distort or constrain academic inquiry. This paper explores how such interventions affect historical research and the core values of academic integrity. As a historian working on sensitive topics such as genocide studies, I have encountered striking examples of GenAI ‘censorship’: terms central to scholarship flagged as inappropriate, digitisation processes prematurely halted, and historical representations altered or withheld. These incidents reveal not only the fragility of GenAI as a research tool but also the risks of allowing opaque corporate logics to filter access to knowledge. This paper adopts an interpretive, case-based approach. Drawing on documented encounters with GenAI systems, I analyse the implications of content moderation for honesty, trust, fairness, and responsibility in academic practice. Rather than presenting restrictions as isolated glitches, I interpret them as symptoms of deeper tensions: between protection and academic freedom, between global North–dominated development practices and diverse scholarly needs, and between efficiency and integrity in research. I argue that humanities perspectives are essential to navigating these tensions. By foregrounding nuance, contextual understanding, and ethical clarity, humanities scholars are uniquely positioned to contribute to GenAI governance and ensure that moderation policies safeguard — rather than undermine — scholarly inquiry.
Presenters
Lornawaddington WaddingtonAssociate Professor, History, University of Leeds, United Kingdom
Details
Presentation Type
Paper Presentation in a Themed Session
Theme
Beyond Borders: The Role of the Humanities in Reimagining Communities
KEYWORDS
GenAI, Humanities, Censorship, Bias
