Insurance Europe has published its response to a consultation by the European Commission on its proposals for a Regulation on artificial intelligence (AI).
Insurance Europe welcomed the overall objective of the Commission to create a proportionate and principles-based horizontal framework of requirements, without unduly constraining or hindering technological development and innovation.
The introduction of harmonised rules on AI, however, requires a very clear and precise definition of an AI system. The Commission has based its definition on the OECD’s definition of an AI system, which it defines as “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments”. While there is currently no universally agreed upon definition of an AI system, insurers believe that the OECD definition is an appropriate basis to use for any European approach, particularly given the inherently global nature of AI systems and the need to ensure consistency at an international level.
The definition of an AI system as currently proposed in Article 3 of the draft Regulation, however, significantly widens the OECD definition by also including software within its scope. This will result in the inclusion in its scope of systems, techniques and approaches that should not be considered as AI and will generally create confusion and a lack of legal certainty.
The broad definition of an AI system in the draft Regulation should therefore be narrowed to be fully aligned with the OECD and avoid the risk of inconsistent and divergent classifications of AI systems.
Insurance Europe also stressed the importance of ensuring consistency with the definitions used in any existing or upcoming European legislative texts that address AI, including in particular the forthcoming liability framework for AI.