Session I-07 - Topics: AI Practices in Language Education; Designing Curriculum-Aligned AI Agents

In Fall 2024, McGill’s French Language Centre created the Committee for Languages, Academic Integrity and AI (CLAI) to support emerging pedagogical needs related to generative AI in language teaching. Bringing together instructors, staff, students, and researchers, the committee develops practical approaches that enhance learning, reduce plagiarism, and uphold academic integrity. This presentation shares effective activities, processes, and community-building strategies, highlighting adaptable practices that foster sustainable collaboration across language programs responding to AI‑driven pedagogical change.

As AI tools increasingly enter classrooms, their effectiveness depends on how well they align with curriculum, pedagogy, and teacher realities. This session presents a practitioner-informed framework for designing customized AI agents grounded in Québec’s MEQ curriculum and real educator feedback. Rather than relying on generic AI chatbots, we build focused, role-specific agents designed for clearly defined teaching purposes. Using survey data, tutor feedback, and classroom consultation, we identify recurring instructional challenges and design agents to address them. Examples include an AI Science Tutor targeting common misconceptions, a Differentiation Agent supporting pedagogical flexibility, and a Preschool Observation Tool aligned with competency-based documentation practices. Our five-phase development cycle integrates needs analysis, curriculum embedding, safeguard design (including Bill 25 compliance), pilot testing, and iterative refinement. Early implementation insights highlight the importance of clearly defined AI roles, transparency about limitations, and strong curriculum anchoring to build teacher trust. Participants will gain a practical, adaptable framework for developing context-sensitive AI tools that strengthen instruction while preserving teacher agency and professional judgment.