Shaping ambidextrous organisations through AI and decision-making: a distinct path for family firms?

Abstract

Purpose Despite increasing research on artificial intelligence (AI) in business, further studies are needed to understand how AI adoption shapes existing and develops new organisational capabilities. This paper aims to examine how AI adoption fosters ambidexterity, both directly and indirectly, through decision-making comprehensiveness (DMC), while also exploring the role of family involvement in this process. Design/methodology/approach We gathered data in a 2-wave survey among 582 management-level participants from UK firms addressed through the Prolific platform. A moderated mediation model was tested in SPSS PROCESS. Findings We find evidence of partial mediation, as AI adoption directly and indirectly fosters ambidexterity through DMC. However, no moderating effect of family involvement is observed. Family firms leverage AI for ambidexterity as effectively as non-family firms, with their focus on long-term survival and adaptability complementing AI-driven decision-making comprehensiveness without compromising core values or socioemotional wealth. Practical implications Managers should consider adopting AI technologies as a strategic enabler to improve DMC and enhance ambidexterity. Our results suggest that family firms may benefit equally from AI despite potential hesitations. We provide suggestions for family firms on how to facilitate AI adoption while overcoming scepticism. Originality/value Our study responds to calls for insights into organisational constructs that clarify the mechanisms behind AI integration and capability development. By examining the role of family involvement, we explore how family businesses can adopt AI to foster innovation capabilities while preserving their legacy. In doing so, we bridge AI research with the family business literature.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 18 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.