AI-DeFi Systemic Risk: A Complex Adaptive System Framework

Abstract

Artificial intelligence (AI) is altering the systemic risk topology of decentralized finance (DeFi). While prior studies examine vulnerabilities in each domain, their intersection remains undertheorized. This paper frames the AI-DeFi ecosystem as a complex adaptive system (CAS) to identify three mechanisms by which AI reshapes systemic risk: algorithmic homogeneity, which synchronizes agent responses; dynamic network rewiring, which amplifies structural fragility; and emergent multi-agent behavior, which produces contagion without centralized intent. Together, these forces introduce endogenous, path-dependent failure modes opaque to reductionist analysis. This framework provides a foundation for empirical analysis and anticipatory governance in AI-mediated financial systems.