The effects of artificial intelligence applications in educational settings: Challenges and strategies

Abstract

With the continuous intervention of AI tools in the education sector, new research is required to evaluate the viability and feasibility of extant AI platforms to inform various pedagogical methods of instruction. The current manuscript explores the cumulative published literature to date in order to evaluate the key challenges that influence the implications of adopting AI models in the Education Sector. The researchers' present works both in favour and against AI-based applications within the Academic milieu. A total of 69 articles from a 618-article population was selected from diverse academic journals between 2018 and 2023. After a careful review of selected articles, the manuscript presents a classification structure based on five distinct dimensions: user, operational, environmental, technological, and ethical challenges. The current review recommends the use of ChatGPT as a complementary teaching-learning aid including the need to afford customized and optimized versions of the tool for the teaching fraternity. The study addresses an important knowledge gap as to how AI models enhance knowledge within educational settings. For instance, the review discusses interalia a range of AI-related effects on learning from the need for creative prompts, training on diverse datasets and genres, incorporation of human input and data confidentiality and elimination of bias. The study concludes by recommending strategic solutions to the emerging challenges identified while summarizing ways to encourage wider adoption of ChatGPT and other AI tools within the education sector. The insights presented in this review can act as a reference for policymakers, teachers, technology experts and stakeholders, and facilitate the means for wider adoption of ChatGPT in the Education sector more generally. Moreover, the review provides an important foundation for future research.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 494 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.