Disruptive innovation ecosystems for sustainability, the circular economy, and artificial intelligence

Abstract

This study examines the structural tensions that hinder the effective integration of sustainability and the circular economy within contemporary innovation ecosystems, with particular emphasis on Latin America. To this end, a critical analysis was conducted of specialized literature published between 2020 and 2025, complemented by bibliometric mapping using Connected Papers and Gephi, and managed in Zotero according to criteria of recency, peer review, and thematic relevance. The results identify three major barriers to advancing toward circular models. First, there are normative, financial, cultural, and technological constraints that operate in a systemic way. Second, artificial intelligence plays an ambivalent role: it can enhance coordination, innovation, and decision-making capacities, while simultaneously generating risks associated with algorithmic bias, power concentration, and governance capture. Finally, collaborative governance mechanisms remain insufficient to articulate heterogeneous actors within fragmented institutional settings. Overall, the findings suggest that the growth in academic production does not automatically translate into tangible progress in circularity, as reflected in the decline of the global circularity index in recent years. For this reason, overcoming these tensions requires innovation ecosystems capable of integrating multilevel governance, blended financing with risk-sharing schemes, and verifiable impact metrics. In this context, artificial intelligence can enhance these capacities only if its adoption is aligned with explicit ethical frameworks, transparency, and institutional structures that ensure an equitable distribution of benefits.