The proposed EU Directives for AI liability leave worrying gaps likely to impact medical AI

Abstract

On 28 September 2022, the European Union (EU) Commission published two new proposals to make liability rules suitable for the digital age, in which Arti??cial Intelligence (AI) systems will have a common place in society. Through the proposed amendments to the Product Liability Directive (PLD)1, and the new AI Liability Directive (AILD)2, the Commission aims to update the laws governing manufacturers strict liability for defective goods (from medical devices to smart technologies). The revised rules aim to provide businesses with greater legal security so they can invest in developing new, cutting-edge products. At the same time the Commission aims to strengthen the position of people to receive just compensation when faulty goods, including digital goods, cause them harm. Both these proposals aim to harmonise and strengthen the product liability regime, ??lling the gap left by the Product Liability Directive adopted in 1985 that currently still regulates AI system liability. Under the current regime, it is dif??cult to get compensation for harm caused by faulty AI systems because of the way the law is phrased. For this reason, the PLD Proposal and the AILD establish two much needed new liability regimes.