Withdrawal of the AI Liability Directive

Contained within Annexes to the European Commission’s Work Program for 2025, was the withdrawal of the proposal for a Directive of the European Parliament and of the Council on adapting non-contractual civil liability rules to artificial intelligence, more commonly referred to as the AI Liability Directive (“AILD”). This withdrawal has been met with mixed response as experts assess the impact that this will have and whether it signals a change in attitude at EU level in relation to the regulation of AI within the Union. 

The AILD was proposed, alongside proposed amendments to the Product Liability Directive relating to AI systems which were introduced in December 2024, to harmonise and update Member State’s non-contractual, fault based rules in respect of claims for injuries caused by AI systems. In a previous article we outlined what was envisaged to be implemented under the AILD in an effort to make it easier for individuals to make civil claims where they suffered damage as a result of AI. This included rules in respect of the provision of evidence by Defendants (providers, operators or users of AI systems) and, importantly, shifting the burden of proof to AI providers by including a presumption of causation between a fault in an AI system and damage suffered to a person - subject to certain conditions. 

The decision to withdraw the AILD, explained in the Annex as being due to “no foreseeable agreement”, comes at a time where much debate is being had in respect of the differing regulatory approaches between the EU and other jurisdictions such as the United States and China with many, including most recently American Vice-President JD Vance, criticising the EU’s policy in respect of AI as being overly regulatory at the expense of innovation. In addition, in January of this year the American Chamber of Commerce to the EU published a position paper in which they called for the withdrawal of the AILD citing what they believed to be “adding unnecessary complexity and uncertainty to Europe’s AI regulatory landscape” which they believed likely to “hinder innovation and disrupt established business practices”. Instead, they argued, regulatory simplification should be pursued. 

However, it is not certain that the AILD, which was proposed in order to ensure accountability, would have hindered innovation. In this regard, it is of note that the AILD would, unlike the EU AI Act for example, only have become relevant where damage, caused by AI, had actually occurred. Secondly, it could be argued that by removing the proposal for a harmonised approach amongst Member States, developers, importers distributors and deployers of AI systems will now have to tackle the varying approaches and legal frameworks of 27 individual Member States in respect of AI Liability rather than one unified approach. This might be an unwelcome task for start-ups and SMEs particularly.  

However, in addition to the general push for regulatory simplification some argue that the combination of the provisions in the AILD and the changes adopted to the Product Liability Directive, where for example, a party in an action for damages may no longer be required to prove negligence where an AI system was defective, would create a hostile legal environment for AI companies.  

It remains to be seen whether this issue will be re-explored by the Commission but it would appear that, for the time being, the Commission’s primary concern is not being left behind in the pursuit of AI innovation and adoption in the Union.   

If you are seeking any legal advice with regards to AI please contact Jon Legorburu, Ben Grogan or any member of our Artificial Intelligence (AI) Group.