High Risk Artificial Intelligence Systems and Legal Doctrine of Essential Facilities: in Search for a Dynamic Model

Abstract

The Regulation of the European Parliament and of the Council on laying down harmonised rules on Artificial Intelligence and amending certain Union Legislative Acts (Artificial Intelligence Act) targets high risk artificial intelligence systems as one of its primary areas of regulatory scope. High risk artificial intelligence systems are considered as software that is developed to use machine learning approaches like supervised, unsupervised and reinforcement learning, deep learning; logic and knowledge-based approaches like knowledge representation, inductive (logic) programming, knowledge bases, inference and deductive engines, symbolic reasoning and expert systems and statistical approaches like Bayesian estimation, search and optimization methods. The essential facilities doctrine in Competition Law / Antitrust Law state that owner(s) of an essential facility for effective competition must provide access to that facility to other competitors in relevant market at a reasonable price. This paper correlates high risk artificial intelligence systems in the scope of Artificial Intelligence Act as potential essential facilities under certain conditions. The paper follows with normative analysis of regulatory requirements of the Artificial Intelligence Act for high risk artificial intelligence systems in light of the essential facilities doctrine. In the final part paper detects primary normative content for future development and outlines dynamic regulatory model for high risk artificial intelligence systems.

Keywords

• artificial intelligence
• regulation
• essential facilities doctrine
• high risk artificial intelligence systems