Prompt engineering as an extension of information literacy: Implications for library instruction programmes

Generative artificial intelligence (AI) is now a routine information tool for students and knowledge workers, and the quality of what they obtain depends in part on how they instruct it. This article argues that prompt engineering is best understood as an extension of information literacy, and that library and information professionals should lead its teaching in partnership with other units. Using conceptual analysis, it maps prompt competencies onto the six frames of the ACRL Framework for Information Literacy for Higher Education and, unlike earlier mappings, identifies where the analogy fails. The mapping holds for question formulation, evaluation and refinement. It fails where a language model generates text and does not retrieve documents, where its sources cannot be inspected, and where it fabricates citations. Those failures are the strongest argument for librarian involvement, because they call for verification skills that prompt technique alone does not supply. The article proposes an instructional design in four strands, including one for workplace settings, sets out how each can be assessed, and considers equity of access to instruction. It also addresses two objections: that prompt engineering is a transient skill, and that librarians are not yet prepared to teach it. The argument is conceptual and requires empirical evaluation.