Use of Large Language Model Tools in Academic Coursework: Opportunities and Challenges in Higher Education

The rapid development of large language model tools has altered the way university students approach academic coursework. These tools can explain difficult concepts, generate and revise text, provide formative feedback, support brainstorming, assist with programming, translate academic material, and facilitate personalised interaction with learning content. Their growing use creates important opportunities for higher education, particularly where students require immediate feedback or where instructors work with large classes and limited time. At the same time, large language models raise substantial concerns regarding academic integrity, authorship, factual accuracy, fabricated references, over-reliance, privacy, bias, unequal access, and the validity of conventional assessment practices. This paper critically examines the opportunities and challenges associated with the use of large language model tools in academic coursework. It adopts an integrative analytical review based primarily on recent peer-reviewed empirical studies, systematic reviews, and international guidance published since the widespread availability of conversational generative artificial intelligence. The available evidence indicates that carefully structured use can improve academic performance, writing, formative feedback, engagement, and some forms of problem solving. However, educational benefits depend strongly on task design, student artificial-intelligence literacy, verification practices, and continued human judgement. Unrestricted substitution of machine-generated work for students' intellectual effort can weaken independent reasoning and undermine the validity of assessment. The paper argues that higher education should move beyond a simple choice between prohibition and unrestricted adoption. Instead, institutions should integrate large language model tools through transparent rules, process-based assessment, disclosure of material use, source verification, oral or reflective components, and explicit instruction in critical artificial-intelligence literacy. The central educational challenge is therefore not whether students will encounter these tools, but how their use can support rather than replace the intellectual processes that academic coursework is intended to develop.