Systematic review of ChatGPT in higher education: Navigating impact on learning, wellbeing, and collaboration

The rapid evolution of artificial intelligence (AI)-based chatbots has significantly influenced higher education in the last couple of years, with large language models (LLMs), such as OpenAI's ChatGPT, emerging as transformative tools. ChatGPT is actively and widely adopted for academic applications, including tutoring, essay drafting, content clarification, and feedback generation. While its integration into educational settings offers numerous advantages, from enhancing learning outcomes and reducing test anxiety to fostering engagement and collaboration, it raises concerns regarding academic integrity, critical thinking, and over-reliance on AI-generated content. This paper presents a systematic review of existing literature on using ChatGPT in higher education, following a PRISMA-based methodology to analyze 39 studies published between 2023 and 2024. The review synthesizes findings across five key dimensions: academic performance, mental health, second-language learners, social influence, and inspiration. The results indicate that while ChatGPT improves accessibility to knowledge and provides cognitive and emotional support, its misuse may hinder independent learning and increase technostress. Furthermore, this review explores ethical and policy considerations surrounding ChatGPT's adoption, emphasizing the need for balanced integration strategies. Key recommendations include institutional guidelines, AI literacy initiatives, and frameworks for ethical integration. This review offers critical insights for educators, policymakers, and researchers seeking to navigate the evolving academic landscape shaped by AI.