Human coaches and AI coaching agents: an exploratory quasi-experimental study of workplace client attitudes

Purpose The aim of this exploratory study was to examine workplace managers’ attitudes to artificial intelligence (AI) coaching agents compared with human coaches, along with the factors influencing AI technology adoption in the workplace. Design/methodology/approach A small-scale quasi-experimental pilot study featured white collar workers (n = 63) and used a custom-designed questionnaire to compare client responses to engaging with a human coach or an AI coaching agent, Alpina, across six factors: developing new insights, working alliance, goal attainment, commitment, trust, confidentiality and shame. Findings Following a single coaching session, coaching clients reported higher scores when working with a human coach than clients working with an AI coaching agent on all factors. Research limitations/implications Caution is needed given the limited sample size, the use of a single session to evaluate and the use of a custom-designed measure. Practical implications Whilst AI coaches continue to improve technically and vary widely in functionality and sophistication, in this study, clients appear to rate human coaches more highly than the featured AI coach on a range of factors, such as insights and goal attainment. Further research is needed to validate these exploratory results and test how speech-to-speech or other factors may influence user ratings. Originality/value This is the first study exploring employee attitudes by comparing human and AI coaches across a range of factors.