An An Ethical Framework for Mitigating AI Algorithmic Bias in Information Resource Development

Abstract

Abstract This opinion paper takes up the crucial topic of algorithmic bias for AI powered collection development, asserting that these biases have the potential to create disparities, stifle intellectual openness and undermine principles in academic libraries. The paper’s novel contributions lies in the development of a library-specific ethical framework within collection development processes rather than treating ethics as abstract principles. By examining algorithmic bias through data selection, model design, and institutional governance, the paper proposes mitigation strategies grounded in library and information science and librarianship. The paper proposes a framework to mitigate algorithmic bias in collection development through transparency in AI modelling, encouraging systematic collection of diverse data inputs, ensuring there are appropriate audit mechanisms developed and creating a culture of ethical AI literacy for library professionals. The paper argues for a forward-looking, user-based approach to using AI in collection development that ssupports rather than undermine intellectual freedom, equity, and inclusivity. Keywords: Algorithm Bias; Artificial Intelligence; Collection Development; Ethical Framework; Information resources; Academic Libraries.

Ähnliche Arbeiten