Empowering, Not Replacing: Using Generative AI to Coach Educators in Providing Effective Feedback

Abstract

As generative AI (GenAI) technologies become increasingly integrated into educational contexts, the importance of supporting rather than replacing educators with AI is gaining prominence. This study introduces Feedback Tutor, a GenAI-powered tool designed to enhance teacher feedback literacy by coaching educators in crafting effective, rubric-aligned feedback, while preserving their agency. It explores how human-AI collaboration can empower educators through a professional development intervention using Feedback Tutor. Eleven educators from two institutions participated in a workshop and follow-up interviews, with qualitative data analyzed thematically. Four key themes emerged: (1) AI-supported professional growth, where educators valued the tool’s coaching capabilities for improving clarity and feedback quality; (2) Human-AI Negotiation and Agency, highlighting the tool’s role as a collaborative partner rather than a replacement; (3) Socio-Technical Integration, emphasizing the importance of aligning AI tools with existing educational workflows and infrastructure; and (4) Ethical and Pedagogical Considerations, underscoring the need for balanced feedback that supports student autonomy and learning. Findings suggest that GenAI tools like Feedback Tutor can significantly enhance educators’ feedback practices and professional growth when designed with pedagogical intent and ethical sensitivity. However, challenges such as lack of content ingestion, and usability constraints must be addressed to optimize adoption and impact. This study contributes to the growing discourse on AI in education by demonstrating how thoughtfully designed AI tools can augment, rather than diminish, the human elements of teaching and learning.