Challenges of regulating and supervising the use of artificial intelligence in the insurance industry

Abstract

The use of AI algorithms raises a large number of interesting regulatory issues in the insurance industry for which there is no comprehensive regulatory framework today. Who bears responsibility for a decision when it is based on an algorithm? How can supervisors monitor a decision-making process when its basis is a neural network that functions as a kind of black box? Insurance companies are already using AI algorithms, and supervisors have realized that they need to quickly find answers to the question of under what conditions they should allow such use. This paper presents the development of AI algorithms for use in the insurance industry, highlights the legal challenges of their use, addresses the positioning of regulators on these legal challenges, and provides an outlook on potential further developments.