Unpacking service failures in artificial intelligence: future research directions

Purpose The present study undertakes an extensive review of the causes of service failures in artificial intelligence (AI) technology literature. Design/methodology/approach A hybrid review has been employed which includes descriptive analysis, and bibliometric analysis with content analysis of the literature approach to synthesizing existing research on a certain topic. The study has followed the SPAR-4-SLR protocol as outlined by Paul et al . (2021). The search period encompasses the progression of service failure in AI from 2001 to 2023. Findings From identified theories, theoretical implications are derived, and thematic maps direct future research on topics such as data mining, smart factories, and among others. The key themes are being proposed incorporates technological elements, ethical deliberations, and cooperative endeavours. Originality/value This research study makes a valuable contribution to understanding and reducing service defects in AI by providing insights that can inform future investigations and practical implementations. Six key future research directions are derived from the thematic and cluster discussions presented in the content analysis.