As AI encompasses a set of technologies that are still at an early stage of development, legislating on liability for such highly advanced systems should be deferred until their specific risk potential can be better understood in the context of their use in different lines of business and the needs of those sectors. Any new rules at EU level would be useful and appropriate only to address any potential gaps where current rules and regulations are found to be insufficient. In our view, this is currently not the case.
Strict liability schemes only work when the risks to be covered are sufficiently similar and when specific market pre-conditions are met (availability of sufficient data, adequate competition, insurers’ interest in providing cover and sufficient reinsurance capacity). This is not the case for AI, which covers a very wide range of different appliances and uses. Without these conditions in place, making product liability insurance mandatory could end up doing more harm than good, at national level and especially at EU level.
Still have questions?
|Contact FFI experts
