ABSTRACT
This chapter discusses selected pivotal legal-methodological challenges to be tackled at the outset of any policy, legal, particularly regulatory and supervisory approach to Artificial intelligence (AI) and machine learning (ML) in the financial sector. It explores the definitional approaches to AI, ML and related phenomena. Any legal-systematic approach to AI and ML requires defining it at the outset. The lack of definitional clarity of the term AI illustrates the so-called AI effect. At European Union (EU) level, the Commission considers ML a type of AI, working by identifying patterns in available data and then applying the knowledge to new data, adding in a footnote, that finding the pattern is itself sometimes the goal of the activity, further explaining the latter. The Commission does not mention the finance sector in the first criterion of a high-risk AI application. At national level, particularly states have adopted strategies on AI.
