Can AI transform public decision-making for sustainable development? An exploration of critical earth system governance questions

Abstract The inability of global governments to meaningfully progress sustainable development over the past three decades is deeply concerning. AI is increasingly framed as a solution for achieving such outcomes, sometimes uncritically. We argue that: 1) for AI to improve public decision-making, the conditions and factors influencing public decisions must be better understood and considered; 2) to mainstream AI-enabled insights, transformations of those conditions and factors are necessary; and, 3) critical governance questions about those transformations must be addressed. To develop our arguments we draw on: original research identifying factors shaping public decision-making; ongoing interdisciplinary research exploring conditions that influence the use of AI for sustainable development policy; and, conceptual framings from literature concerned with transitions, earth system governance, leverage points and policy entrepreneurship - all sharing ambition to understand transformative change. In so doing, we seek to advance critical knowledge on the, potentially, transformative implications of AI in public decision-making.