A framework for quality assurance of artificial intelligence in archives and records management in Eswatini

Résumé fourni par la source

Limited research in Eswatini’s information science, archives, and records management (ARM) sectors has created gaps in understanding the implications of integrating artificial intelligence (AI). This study explores the need for a quality assurance (QA) framework for AI in ARM in Eswatini. The study adopts an interpretive research paradigm and conducts a literature review using databases such as EBSCO, Scopus, and Google Scholar, covering materials published between 2009 and 2025. Forty-one relevant sources, including journal articles, reports, and monographs, were analyzed. The findings indicate that, although there is growing interest in applying AI in ARM, stakeholders remain concerned about ensuring the quality, transparency, and reliability of such systems. This study lays the foundation for future empirical research. It proposes a framework that includes a national policy, ethical algorithm development, robust data governance, rigorous validation, staff training, and continuous monitoring. These recommendations aim to promote ethical compliance, ensure data integrity, and support sustainable AI practices.