Evaluating ChatGPT for Library Metadata Creation: Accuracy, Consistency, and Challenges

In recent years, Artificial Intelligence (AI) has become an essential tool in various fields, including Library and Information Centres (LICs). It is very important for library services to create metadata, which is structured information that describes, explains, locates, or otherwise makes it easier to find, use, or manage an information resource. This study was conducted to identify how accurate and consistent ChatGPT provides metadata for books and also identified the challenges associated with using ChatGPT. To achieve the objectives, The ChatGPT 4o open access part was used. Three books were used to craft queries that prompted ChatGPT to provide metadata as well as MARC records for the books. It was found that ChatGPT provided accurate and consistent metadata and MARC records for the information resources identified; it gives more metadata elements of books that have e-versions online. The challenges associated with using ChatGPT are inefficiency in a fluctuating network environment and its inherent bias to books that are not present online. It was recommended, among other things that librarians should not depend absolutely on using ChatGPT in cataloguing and classification of information resources; ChatGPT should be used in a strong high bandwidth network environment and; a separate query should be made when in need of call number for books.