Since the advent of ChatGPT as a revolutionary tool for generating human-like text, researchers have increasingly focused on detecting AI-generated text to address the rising concerns about the authenticity and potential misuse of AI-generated text. While previous studies have mainly focused on distinguishing between AI-generated text and human-written text, our research aims to further explore identifying different uses of a generative model. Consider a student assignment that incorporates AI only to refine writing quality. Existing detection tools often label such assignments as entirely AI-generated, raising doubts about the student’s authenticity. To address this, we propose AICIS (AI Contribution Identification System), a multi-class classification system, that classifies text into human-written, AI-enhanced (human-written text polished by AI), or AI-originated (fully generated by AI). Our experimental results demonstrate AICIS’s ability to achieve an overall accuracy of 97.9% and an impressive precision of 99.47% for AI-originated text. Additionally, we compared AICIS against GPTZero, a leading AI detection tool, revealing GPTZero’s tendency to misclassify AI-enhanced samples as AI-only. By addressing this limitation, AICIS offers a promising solution to preserve text authenticity in various domains, such as education and academia.