Day 1

6/26/2025

10:45 - 12:00

Station C: Speaking to a Machine

AI chatbots are increasingly being used to support speaking practice in higher education—offering students low-stakes environments to improve fluency, pronunciation, and turn-taking. However, as institutions move toward standardizing these tools, important ethical questions emerge:

What data is being collected during AI conversations?

Can we ensure fairness and equal access to high-quality interaction?

How do we prevent over-reliance on chatbots that lack real communicative feedback?

This presentation explores the ethical dimensions of integrating AI chatbots (e.g., ChatGPT, AI avatars) into speaking practice in university classrooms. Drawing from classroom experiments and teacher feedback at FSMVU, the presenter will share:

Examples of chatbot-based speaking tasks and what students found helpful or problematic Risks of bias, miscommunication, and hallucination in AI-led interaction Institutional responsibilities when students are encouraged (or required) to engage with generative tools. The session will also offer practical recommendations for ethical standardization, such as: Setting transparency guidelines (e.g., disclosing bot limitations and data use) Designing chatbot tasks that foster reflection rather than passive interaction Using human-led debriefing or peer discussion to balance AI conversations

Learning Objectives:

1. Identify ethical risks involved in using chatbots for speaking activities

2. Explore how to structure AI conversations to support pedagogy without replacing authentic interaction

3. Consider policy-level decisions around AI use in oral practice

Expected Outcomes:

Attendees will leave with sample speaking activity designs, a checklist for ethical implementation, and ideas for building consistent policy around chatbot integration in speaking courses.