INTEGRATING RESPONSIBLE ARTIFICIAL INTELLIGENCE (AI) IN CEFR B2 ESL ACADEMIC READING AND WRITING: A FRAMEWORK FOR AFFECTIVE-BASED TASK DESIGN

As Generative Artificial Intelligence (GenAI) becomes deeply embedded in higher education, the focus of ESL pedagogy must shift from mere technical proficiency to the "responsible use" of these tools. This conceptual paper provides a comprehensive overview and a structured guide for integrating AI into academic reading and writing tasks specifically for learners at the CEFR B2 level. Unlike traditional approaches that prioritise cognitive output, this framework emphasises the affective domain—focusing on student values, ethical awareness, motivation, and autonomy. The paper proposes a multi-stage integration model where AI serves as a "study partner" during topic exploration, annotation, and infographic development. Central to this model is the consultation and reflection table, a metacognitive tool designed to foster academic integrity and self-regulation. Furthermore, the paper introduces an Outcome-Based Education (OBE) aligned affective rubric that evaluates students across five dimensions: responsibility, integrity, reflection, autonomy, and engagement. By shifting the assessment lens toward the process of human-AI collaboration rather than just the final product, this framework provides educators with practical strategies to mitigate over-dependence and plagiarism. Ultimately, this paper argues that fostering AI literacy through affective-based tasks empowers B2 learners to maintain their unique voice and critical agency in an increasingly automated academic landscape.