Paired Human and AI-Generated Indonesian News Text Dataset (Kompas, Liputan6, Detik; GPT, Gemini, Claude)

This dataset contains 1,640 paired Indonesian news documents (820 human-written, 820 AI-generated) used to evaluate AI-generated text detection. Human articles were collected from three Indonesian online news publishers (Kompas, Liputan6, and Detik), restricted to articles published before 2020 to ensure they predate widespread LLM availability. For each human article, a paired AI-generated article was produced using the same headline and topic across three generative models: GPT (274 documents), Gemini (273), and Claude (273). Each record includes the document id, source publisher, title, URL, publication date, full text, class label (human/AI), and generator. This dataset was collected as part of a study on multi-feature fusion for AI-generated text detection in Indonesian news, published in the International Journal on Informatics Visualization (JOIV).