SOCIETÀ ITALIANA DI DIRITTO ED ECONOMIA
Alessio Azzutti (University of Glasgow)
Abstract
The frenzied race toward Artificial Intelligence (AI) adoption is causing profound transformations within the financial sector, rendering capital markets an increasingly complex system. These dramatic and sweeping changes are most pronounced in data-intensive and high-performance computing domains, such as algorithmic trading. While AI-powered trading offers numerous benefits to financial firms, markets, and society, it also raises significant concerns regarding potential risks to market quality, integrity, and stability. Recent studies underscore the dangers posed by AI advancements, particularly when not accompanied by robust governance and regulatory frameworks, which could lead to new and heightened risks of market abuse. Amidst this risk-prone environment, there is growing recognition among policymakers and financial regulators of the pressing need to regulate AI deployment. This emerging awareness is crucial, as effective AI governance is essential to ensure that the benefits of technological innovation are not overshadowed by its inherent risks. In this very direction, the EU AI Act stands out as a landmark effort in establishing comprehensive AI regulation. Hence, this Article critically examines this fundamental piece of (global) legislation and compares it to sectoral regulation on algorithmic trading. By focusing on key legal provisions, the analysis demonstrates the potential superiority of the EU AI Act’s regulatory requirements for providers of “high-risk” AI systems over those for deployers of algorithmic trading systems under MiFID II. The Article concludes with some ideas for future risk-based regulation of AI applications in financial trading.