Abstract

Recent advances in artificial intelligence (AI) have transformed academic publishing by providing automated tools for journal classification and quality assessment. However, these tools are not immune to errors. This article examines a significant issue where an AI tool erroneously categorizes legitimate sports science journals as predatory. The study explores the underlying causes of this misclassification, including limitations in data quality, algorithmic biases, and fieldspecific nuances. It highlights the impact on researchers and institutions, emphasizing the need for improved AI systems and human oversight. The findings underscore the importance of refining AI methodologies to enhance accuracy and ensure that scholarly publishing systems effectively support academic integrity and research dissemination.