Abstract
As Artificial Intelligence (AI) transforms organisations, the challenges of algorithmic bias and the need for inclusive innovation are paramount. Yet, research often treats these as separate technical problems, overlooking the deep-seated organisational conditions that link them. This paper argues that responsible AI is fundamentally a cultural phenomenon. We develop a multi-level conceptual framework that integrates organisational culture theory with AI ethics and inclusive innovation studies. The framework explains how cultural layers—from visible artifacts and espoused values to deeply embedded underlying assumptions—jointly shape AI development and its outcomes. We theorise that algorithmic bias emerges from culturally embedded practices, while inclusive innovation depends on the alignment of these cultural layers toward equity and reflexivity. By shifting focus from purely technical solutions to the organisational meaning systems that make bias permissible or invisible, this work provides a new foundation for research and practice aimed at cultivating fair, accountable, and inclusive sociotechnical systems.