AI is an approach to organizational change which is based on strengths rather than weaknesses, on visions of what is possible rather than what is not possible. It identifies achievements and “best memories”, and through this technique, locates “where the energy is” in an organization. As a mode of inquiry, AI seems to take better care of informants and participants in research, enabling them to dwell on the best of what is as well as the worst. As a mode of transformation, AI locates the key sources of energy and satisfaction in an organization, and brings out hidden skills, enthusiasms and creativity in staff (and to some extent, in prisoners). There are some links, in terms of underlying principles, between AI and restorative justice. Both approaches achieve positive change because where outcasting and shaming make things worse, limited disapproval, within framework of respect, forgiveness and trust, works better. Both mechanisms depend on generating positive affect. It is concluded that AI constitutes a fair and inclusive research approach that generates a rich and faithful account of a prison to emerge. It generates energy among prison staff that can be harnessed in the direction of better practice. The mechanism at work is a normative process, which seems to engage the research participants in meaningful, constructive and ethically relevant dialogue about their practices and experiences. The special and complex moral environment of the prison makes AI especially relevant. 18 notes, 38 references