Investigating the use and impact of LLM-based AI tools on classroom practices.

The rapid adoption of Artificial Intelligence (AI), especially Large Language Models (LLMs),has changed educational practices by providing teachers with an adaptable tool that canassist them in lesson planning, content generation, assessment marking, and administrationtasks among other tasks. While the potential benefits are widely acknowledged, such asspeeding up administrative tasks and lesson planning, there are concerns regarding ethicaluse, data privacy, misinformation, and adequacy of teacher training. Existing research hasmostly focused on student use of AI, leaving a gap in understanding how teachers use LLMbasedAI in their classroom practices.This study aims to investigate how teachers use LLM-based AI tools before, during, andafter a lesson and how they acquire the knowledge to use those tools effectively andresponsibly. This research will be guided by the Technology Acceptance Model (TAM) andConstructivist Learning (CL) theories and use data from publicly available online forums andother relative studies in a Grounded Theory (GT) approach. This is due to the constraint ofnot being allowed to use primary data and a further limitation to secondary data in terms ofnot gaining access to meeting records of teachers discussing AI that can provide vastlybetter practical experiences than other sources.These findings are expected to contribute to the body of knowledge on AI in education byproviding insights into teachers’ experiences and use of LLM-based AI tools. It can alsoprovide better context for AI governance and teacher support.Key words: Artificial Intelligence, Large Language Model, Technology Acceptance Model,Constructivist Learning, Classroom Practices, Teachers.