The rapid growth of digital content and the increasing complexity of information management have made metadata essential for organizing and accessing digital library collections. Traditional manual methods of metadata creation and maintenance struggle to handle large and diverse collections efficiently. As a result, many libraries are exploring the use of Artificial Intelligence (AI) to automate and improve metadata workflows. This study conducts a systematic literature review to examine how AI is being applied to metadata management in digital libraries, focusing on its benefits, challenges, and future potential. The review follows the PRISMA guidelines and analyzes 24 peer-reviewed studies published between 2005 and 2025. The findings show that AI technologies, particularly Natural Language Processing (NLP) and machine learning, are being used to automate tasks such as metadata generation, classification, and enrichment. Tools like BERT and GPT-4 demonstrate strong performance in tasks like keyword extraction and entity recognition, while hybrid systems combining AI with human oversight help ensure accuracy. Despite these advancements, challenges remain. Many libraries face difficulties integrating AI with existing systems, and concerns about bias, transparency, and data privacy persist. The review also identifies gaps in research, such as the need for more studies on multilingual metadata and real-world implementation. This study provides a comprehensive overview of AI applications in metadata management and offers recommendations for libraries considering AI adoption. It highlights the need for ethical guidelines, staff training, and further research to ensure AI is used effectively and responsibly in digital libraries.
Anastasia Palaska (Wed,) studied this question.