Artificial Intelligence (AI) tools are now deeply embedded in higher education, influencing the ways students plan, draft, and refine their academic writing. Applications such as QuillBot, Grammarly, and Jenni AI have become particularly popular for enhancing accuracy, fluency, and overall productivity. Yet, despite their rapid adoption, evidence remains limited on how satisfied students actually feel when using these tools, especially concerning their perceived usefulness, ease of use, and ethical implications. This review seeks to answer the question: What are the levels and determinants of students’ satisfaction with AI writing tools in higher education? A systematic literature review was conducted following the methodological framework proposed by Shaffril et al. (2021) and guided by the PRISMA 2020 reporting standards to ensure transparency and reproducibility. Searches were performed across Scopus, Web of Science, and Google Scholar in October 2025. Only empirical, peer-reviewed studies published in English between 2020 and 2025 were included, while conceptual and non-empirical works were excluded. Each selected study underwent a quality appraisal to assess methodological rigor and relevance. Data were extracted through a structured coding matrix, and the synthesis was carried out using thematic content analysis. Twenty studies met the inclusion criteria. Across these, three consistent factors emerged as key determinants of satisfaction: perceived usefulness, ease of use, and ethical awareness. Most students reported high satisfaction particularly in relation to improved writing accuracy and confidence though concerns about plagiarism and excessive reliance on automation persisted. The findings suggest a growing need to cultivate AI literacy and ethical awareness to support responsible engagement with AI tools in academic contexts. Nonetheless, the body of evidence is constrained by the predominance of self-reported data, which may limit broader generalisation.