Following yesterday’s plenary vote in the European Parliament to approve its legal affairs committee reports on a civil liability regime for artificial intelligence (AI) and a framework of ethical aspects of AI, robotics and related technologies, Insurance Europe’s director general, Michaela Koller, said:

“We believe it is premature for the Parliament to consider mandatory insurance for high-risk AI systems for two reasons. Firstly, for compulsory insurance to work, there must be sufficient data on the risks. For example, in mandatory motor insurance there is claims data going back decades that can be used to underwrite the risk. However, because high-risk AI systems are still in development, an equivalent volume of data is not yet available. Secondly, risks covered under mandatory policies must be similar enough in nature to be covered under a one-size-fits-all system. However, because AI is used in such diverse ways, the risks it poses in various situations are completely different. Therefore, it is difficult to see how a one-size-fits-all approach could be considered that would be able to cover such diverse risks adequately. Against this background, we take the view that the Parliament’s call for mandatory insurance for high-risk AI would not work in practice.”

Regarding the report on a framework of ethical aspects of AI, Koller commented:

“Existing horizontal and sectoral EU legislation that applies to insurance – including the equal treatment directives, the General Data Protection Regulation, the Solvency II Directive, the Insurance Distribution Directive, the Packaged Retail and Insurance-based Investment Products Regulation and the Distance Marketing Directive for Financial Services – already address many of the potential risks and challenges related to AI. Therefore, additions and overlaps must be avoided, and new rules should only be introduced if gaps are found in existing legislation.”