Bringing the Product Liability Directive into the Digital Age
January 10, 2025
Written by Juliette Mignot | LL.M., CIPP/E
The rapid evolution of technology, particularly in the realms of artificial intelligence (“AI”) and connected systems, has significantly transformed industries and everyday life. This transformation has also introduced new challenges in ensuring consumer safety and accountability. In response, the European Union (“EU”) has revised its Product Liability Directive (“PLD”) and introduced the AI Act, seeking to modernise its legal framework. While these measures represent progress, critical gaps remain, especially regarding liability exemptions for defects beyond a manufacturer’s control.
The revised PLD and the AI Act serve complementary roles in managing risks associated with AI and digital products. The AI Act focuses on prevention, establishing comprehensive rules to ensure AI systems are designed and used safely. It introduces a risk-based classification of AI systems, imposing stringent requirements on high-risk applications, such as those used in critical infrastructure, healthcare, or recruitment. By setting transparency, documentation, and oversight obligations, the AI Act aims to mitigate risks at the outset – thereby preventing harm.
Conversely, the PLD provides a mechanism for compensation when these safeguards fail. It emphasizes strict liability for manufacturers, covering material losses, personal injuries, and even data loss caused by defective products.
Together, the revised PLD and the AI Act aim to ensure accountability and trust in AI technologies while fostering innovation.
These distinctions ensure that victims can seek redress regardless of the type of harm or responsible party.
Article 11 of the PLD allows manufacturers to be exempt from liability if a defect arises outside their control. This provision, designed to balance accountability with innovation, raises significant concerns in the context of AI.
For instance, in complex digital systems, defects may emerge from third-party software updates or interactions with other connected devices. By providing that manufacturers would not be liable in such scenarios, the directive potentially undermines consumer protection. The “black box” nature of AI, where decision-making processes are opaque, exacerbates this issue by making it challenging to trace the origin of defects.
Critics argue that this exemption shifts the burden onto consumers, who may lack the technical expertise to identify the true cause of harm. Moreover, it creates a regulatory loophole that could discourage manufacturers from implementing robust post-market monitoring and collaboration with third-party service providers.
To address these shortcomings, the PLD could adopt stricter requirements for manufacturers to maintain oversight of their products throughout their lifecycle. This includes mandating collaboration with software providers and ensuring compatibility with connected devices. The directive could also introduce a shared liability framework to hold all contributing parties accountable, promoting a culture of safety and transparency.
Additionally, aligning the PLD more closely with the AI Act’s documentation and monitoring requirements could reduce ambiguities. Clearer guidelines on manufacturers’ responsibilities for third-party updates and interactions would enhance consumer trust without necessarily stifling innovation.
The revised Product Liability Directive and the AI Act represent significant steps toward a modern regulatory framework. However, the liability exemptions in Article 11 of the PLD threatens to undermine the directive’s consumer protection goals. Striking a fairer balance between fostering innovation and ensuring accountability is essential. By addressing these issues, the EU can set a global standard for regulating technology in the digital age.
Farida Arhab-Girardin, 'Mise sur le marché et régimes de responsabilité: adaptation, innovations et difficultés probatoires' RDSS 2024 776.
Eugénie Petitprez and Rodolphe Bigot, 'Premières vues sur la directive européenne (UE) 2024-2853 relative à la responsabilité du fait des produits défectueux' Dalloz actualité (28 November 2024).
European Parliament, 'EU AI Act: First Regulation on Artificial Intelligence' Topics (1 June 2023 – last updated 18 June 2024)
https://www.europarl.europa.eu/topics/en/article/20230601STO93804/eu-ai-act-first-regulation-on-artificial-intelligence accessed 6 January 2025.
