Purpose:
This study explores how postgraduate coaching students experience generative AI (specifically ChatGPT) as a simulated coachee and feedback provider. It examines how such interactions support or hinder the development of coaching competence and reflective practice, drawing on Kolb’s experiential learning theory (ELT) and Bandura’s social cognitive theory (SCT).
Design/methodology/approach:
Twelve postgraduate students each conducted a simulated coaching session with ChatGPT in the coachee role, received AI-generated feedback, and submitted written reflections. A follow-up focus group facilitated collective sense-making. Data were analysed using reflexive thematic analysis, informed by ELT’s learning cycle and SCT’s emphasis on self-efficacy, feedback, and personal agency in learning.
Findings:
Five themes were developed: (1) Authenticity and Engagement, (2) Cognitive Load and Processing Challenges, (3) Perceptions of Feedback and Learning, (4) Ethical Concerns, and (5) AI’s Role in Coaching Education. Students appreciated the low-stakes practice space and affirming feedback but noted limitations in emotional realism and relational depth. These gaps, along with cognitive strain, ethical concerns, and uncertainty about feedback credibility, shaped how students engaged with the tool. ELT and SCT offered a dual lens to understand how AI mediates reflection and skill development.
Originality:
This study offers theoretically grounded insights into how AI can scaffold reflective learning, self-efficacy, and experimentation, while also surfacing relational, cognitive, and ethical tensions. The findings support a critically informed integration of AI into coaching curricula, underpinned by AI literacy, digital trust, and reflective judgement.