Can AI Trigger the Next Financial Crisis? Algorithmic Trading Black Boxes and the Responsibility Gap

With the rapid advancement of artificial intelligence technologies, algorithm-basedinvestment systems have been spreading rapidly in financial markets. High-frequencytrading, automated portfolio management, and machine-learning-based investmentstrategies are fundamentally transforming the structure of investment decision-making.This transformation goes beyond simple technological innovation and is reshaping boththe judgment structure and the responsibility structure of financial markets.1This study begins with the following questions: Can AI-driven investment systemsbecome the trigger of the next financial crisis? If such a situation occurs, who shouldbe held responsible?This paper introduces the concept of a structural responsibility gap that emergesin algorithmic investment environments. In automated investment systems, data collection, model analysis, signal generation, and order execution operate within a unifiedalgorithmic structure. Investment decisions are therefore performed internally withinthe system without direct human intervention, while responsibility becomes distributedacross developers, financial institutions, investors, and regulatory authorities. As a result, a structural problem arises in which the attribution of responsibility becomesunclear in times of crisis.Thestudy also analyzes how algorithmic synchronization and feedback loops—whichoccur when multiple algorithmic trading systems simultaneously respond to sharedmarket signals—can amplify market volatility and expand systemic risk. Such dynamics provide an important mechanism for explaining sudden market instabilities such asflash crashes.To address this issue, the paper proposes a Hierarchical Joint Responsibility Model.This model restructures responsibility within algorithmic investment environments intothree layers: the design layer, the operational layer, and the authorization layer. Bydoing so, it presents an institutional framework capable of maintaining both traceabilityand normative control of responsibility within automated financial systems.This study emphasizes that AI-based financial systems are not merely technological tools but socio-technical structures that transform the very architecture of financial judgment. The study ultimately aims to provide philosophical and institutionalfoundations for redesigning responsibility structures in increasingly automated financial markets, while also offering conceptual guidance for regulators and policy makersseeking to address accountability challenges in AI-driven financial systems.