Responsible AI Governance as an Organisational Capability: A Theory-Building Integrative Review for Sustainable Organisations

Artificial intelligence (AI) is increasingly embedded in organisational decision-making, creating opportunities for innovation and sustainability while intensifying concerns about accountability, transparency, stakeholder responsibility and human oversight. This structured theory-building integrative review synthesises a purposively bounded corpus of 35 unique works on responsible AI, AI governance, responsible leadership and organisational sustainability. Responsible AI Governance (RAIG) is conceptualised as a multidimensional organisational capability: the repeatable, coordinated and adaptive capacity through which responsible AI principles are translated into governance structures, decision processes, accountability arrangements, stakeholder relationships, human oversight and assurance across the AI lifecycle. Six interdependent dimensions are specified: ethical-strategic alignment, accountability architecture, human oversight and agency, stakeholder inclusion, transparency and assurance and lifecycle governance and adaptability. Responsible leadership is positioned as an antecedent managerial microfoundation rather than a constitutive dimension of RAIG. The framework separates sustainability from the content of RAIG and theorises sustainable organisational and stakeholder outcomes as consequences of effective governance. Propositions and a staged research programme establish a foundation for construct operationalisation, cross-context validation and subsequent nomological testing.