Using AI in the Pursuit of Justice : Balancing Tensions Between Achieving Social Justice Missions and Using Artificial Intelligence for Information Management

Sammanfattning

This paper is a hermeneutic study of the tensions staff at organisations with ethical missions experience and anticipate when using AI tools for information management. It focusses on the contradictions between the desire to use AI tools to maximise productivity to meet their goals and the responsibility to avoid contributing to injustices that damage their missions. The aim is to discover what kinds of AI policies global justice organisations can design to navigate these tensions while maintaining staff consent. A mixed quantitative and qualitative method is used, with surveys and interviews held in parallel, and priority given to thematic analysis of the interviews. Six themes are identified as interpretative repertoires commonly deployed in information workers’ discourse to make sense of the role of AI tools in promoting or hindering their organisations’ missions. Each theme suggests a different kind of policy response. Internal tensions related to the use of specific tools, whether framed as dilemmas, dialectics or paradoxes, appear more tractable than external tensions between competing overarching sociotechnical visions. Policymaking processes are found to be cautious and experimental, balancing various concerns beyond either productivity or ethical purity. It recommends policymaking should address all six themes to gain broadest consent.