The promise and challenges of machine learning in finance

Natalia Bailey

Data and data-intensive technologies such as artificial intelligence (AI) and machine learning (ML)¹¹ AI is the capacity for machines to resemble human intellectual abilities. It is a broad field with many sub-fields and related fields, including machine learning. ML, a component of AI, provides systems with the ability to automatically learn over time, generally from the large quantities of data. are at the core of the digital transformation of the financial services industry. As there are a variety of AI/ML techniques with various levels of sophistication and opacity, there is no single way of identifying and managing risks related to AI or ML. Rather, there is a wide array of techniques and solutions to managing risks associated with ML models.

Risks associated with ML models are often linked to financial institutions’ model risk management frameworks, and thus the risk controls and approaches taken should be commensurate with the materiality, risk or specific use case. Having strict and rigorous risk controls, and being able to demonstrate competence in the techniques that are being used, are crucial to preserving trust with customers.

Supervisors will assess whether