This chapter assesses whether and to what extent the revised Product Liability Directive (RPLD) provide the manufacturers and providers of AI-powered devices for personalized medical solutions with desirable safety incentives. To this end, the chapter outlines relevant aspects of the current safety regulatory framework including the Medical Device Regulation and the AI Act. In complement, the chapter includes an analysis on the role of the RPLD. This analysis focuses on a hypothetical scenario where an autonomous AI-powered medical device that initially seems safe continues to learn and later begins to provide flawed information resulting in patient harm. As such, the discussion centers on the potential benefits and limitations of the new legislation including the exemptions for ‘later risk’ and ‘development risk’. I argue that the RPLD could incentivize manufacturers to react to any emerging risks that become known or knowable. However, the situation is more nuanced for the first victims of unknown risks. This is because the revised product liability regime, while limiting the opportunity for manufacturers to rely on the ‘later risk’ exemption, lends itself to inconsistent incentives through the ‘development risk’ defense. This raises the question of whether strict liability should be considered for these kinds of products. The legal systems in which fault-based, product, and strict liability regimes coexist are well placed as ‘living laboratories’ where one can readily test how these different liability regimes operate.