From AI potential to sustainable value: The AI-SDGs value cycle as an integrative framework for sustainable innovation

Artificial intelligence (AI) is increasingly positioned as a means of advancing the Sustainable Development Goals (SDGs), yet its contribution remains conditional, uneven, and context-dependent. To explain this variation, we organize the evidence through a single integrative lens: the AI-SDGs value cycle. The cycle comprises five interdependent functions - mapping opportunities, mobilizing capabilities, deploying applications, governing legitimacy and accountability, and evaluating sustainable value creation and destruction. It connects evidence across technologies, teams, organizations, sectors, cities, markets, and public institutions without reducing diverse contributions to a catalog of applications. The synthesis shows that sustainable outcomes arise not from technological capability alone, but from its combination with organizational readiness, sectoral fit, responsible governance, stakeholder representation, and attention to cross-goal and lifecycle consequences. Collectively, the papers move research beyond broad claims about AI's potential toward explanations of when, how, for whom, and at what cost AI supports sustainable development. They establish governance as constitutive of sustainable innovation and reveal value creation and destruction as simultaneous possibilities rather than separable outcomes. Drawing these insights together, this editorial synthesizes the 17 accepted articles in the Technovation Special Issue “Artificial Intelligence for the Sustainable Development Goals” and clarifies their collective contribution. It concludes with a focused research agenda organized around the same value cycle and directed toward cumulative, comparative, and impact-oriented inquiry.