AI in sustainable development management for international trade: opportunities, threats and the growth paradox

Purpose: This study examines mechanisms through which AI supports and undermines sustainable development management in international trade, identifies boundary conditions determining outcomes, and develops management design principles for maximising benefits while mitigating paradoxes and unintended consequences. Methodology: Literature review with mechanism analysis integrating rebound effect theory, employing desk research and deductive reasoning. Findings: Analysis reveals three positive mechanisms (optimisation in bounded systems, enhanced transparency, substitution) and four negative mechanisms (Jevons Paradox, AI infrastructure footprint, social trade-offs, information paradox). Seven boundary conditions determine which of five AI-sustainability relationship patterns emerges: genuine win-win, efficiency paradox, information paradox, trade-off effects, or net negative outcomes. Conditions favouring paradoxical outcomes appear more prevalent than those enabling genuine benefits. Research limitations: Conceptual analysis without empirical measurement of management implementations; literature exhibits publication bias. Practical implications: AI supports sustainability only with absolute regulatory caps, comprehensive system boundaries, and multi-objective optimisation. Intensity-based targets enable paradoxical outcomes. Originality: First application of rebound effect theory to AI in sustainable development management, providing a decision framework for practitioners.