Practical and Ethical Challenges of Large Language Models in Education: A Systematic Literature Review

Abstract

Educational technology innovations that have been developed based on large language models (LLMs) have shown the potential to automate the laborious process of generating and analysing textual content. While various innovations have been developed to automate a range of educational tasks (e.g., question generation, feedback provision, and essay grading), there are concerns regarding the practicality and ethicality of these innovations. Such concerns may hinder future research and the adoption of LLMs-based innovations in authentic educational contexts. To address this, we conducted a systematic literature review of 118 peer-reviewed papers published since 2017 to pinpoint the current state of research on using LLMs to automate and support educational tasks. The practical and ethical challenges of LLMs-based innovations were also identified by assessing their technological readiness, model performance, replicability, system transparency, privacy, equality, and beneficence. The findings were summarised into three recommendations for future studies, including updating existing innovations with state-of-the-art models (e.g., GPT-3), embracing the initiative of open-sourcing models/systems, and adopting a human-centred approach throughout the developmental process. These recommendations could support future research to develop practical and ethical innovations for supporting diverse educational tasks and benefiting students, teachers, and institutions.

Active Inference Goes to School. The Importance of Active Learning in the Age of Large Language Models

“…Manufactured learning environments which leverage exploratory behaviours and facilitate manageable prediction errors are then fundamental when we turn to imagine the implementation of different, by definition disembodied artefacts such as LLMs. LLMs' potential impact on education has generated a variety of responses, from fear to enthusiasm to immediate applicative attempts in a variety of contexts (e.g., Williamson et al, 2023;Devlin, 2023;Milano et al, 2023;Yan et al, 2023;Prihar et al, 2023;Dai et al, 2023;Sanusi et al, 2023; for a summarising commentary, see: Kasneci et al, 2023). If it seems clear that LLMs are here to stay and their integration in education is not challenge-free, then "a clear pedagogical approach with a strong focus on critical thinking and strategies for fact checking are required" (Kasneci et al, 2023.…”

Large Language Models: A Comprehensive Survey of its Applications, Challenges, Limitations, and Future Prospects

“…And what are the associated risks of applying these models in an educational context? It's crucial to initiate the discussion around these models with ethical considerations [339].…”

“How does this Move us Forward?”

“…Some work has focused specifically on chatbots and large language models (LLMs), which are especially relevant to ChatGPT. A recent meta-analysis suggests that AI chatbots can improve student learning outcomes, particularly in higher education (Wu & Yu, 2023); and a systematic review of LLMs in education shows that they can be useful for teachers in several tasks (generating questions, providing feedback, grading essays), but that these capabilities all raise practical and ethical concerns (Yan et al, 2023). Focusing specifically on ChatGPT for education, Su and Yang (2023) describe a number of benefits the tool may have, such as offering personalized and engaging learning experiences for students, and offering suggestions and efficient question-answering for teachers.…”