HUMAN-LANGUAGE MEDIATION IN AI SYSTEMS: A SIX-STAGE PROMPTING FRAMEWORK FOR INCLUSIVE TEFL PRACTICE
HUMAN-LANGUAGE MEDIATION IN AI SYSTEMS: A SIX-STAGE PROMPTING FRAMEWORK FOR INCLUSIVE TEFL PRACTICE

Summary/Abstract: This paper presents a structured six-stage framework for prompt engineering in the context of Teaching English as a Foreign Language (TEFL), with a focus on producing inclusive, culturally sensitive, and pedagogically relevant AI-generated materials. Large Language Models (LLMs) are increasingly used to support language teaching, yet prompt design often remains informal and inconsistent, leading to variable quality and occasional cultural bias. Building on insights from natural language processing, cognitive science, and human–computer interaction, the proposed workflow moves from goal definition through role assignment, contextual embedding, and structured instruction, to iterative refinement and explicit ethical checks. Case studies from TEFL practice illustrate how this approach transforms vague or biased prompts into targeted, level-appropriate, and inclusive teaching resources. By integrating bias-awareness into the prompt itself, the framework ensures that AI outputs align with both pedagogical objectives and ethical standards. This study positions prompt engineering as an essential skill for educators aiming to harness AI as a reliable and equitable teaching partner.

• Details
• Contents