AI systems and non-contractual liability: A European Private International law analysis

Benedetta Cappiello from the University of Milan has recently published a book on European private international law and non-contractual liability for AI systems (AI Systems and Non-contractual Liability: A European Private International Law Analysis, Giappichelli 2022: https://www.giappichelli.it/media/catalog/product/excerpt/9788892143289.pdf). She has kindly provided us with the following abstract:

The advent of AI-systems has fundamentally altered the whole of society and is about to change our daily lives as well as relationships between private parties.

The current challenge for the legislator is to determine a clear legal framework able to firstly, guarantee continued technological development and secondly, to be integrated with already binding sources of law. Whether the said framework will correspond to an already existing one, adapted to AI-systems, or whether it will be an ad hoc framework is still to be scrutinized. What is certain is that the challenge to determine a legal framework assumes a cross-border connotation: only common and shared choices at the supranational level will guarantee the definition of a coherent and effective discipline.

Within the said framework, the present book focuses on the non-contractual obligations which arise within the European Union out of the development and use of AI-systems; more precisely, as for the civil liability regime the advent of AI is about to lead to a paradigm shift in the allocation of liability throughout the “production chain”. Namely, the question has become how to ascertain who is liable for what; the opacity of AI-systems – especially those engaging with machine learning techniques – can make it extremely difficult to identify who is in control and therefore responsible.

Both EU substantive and private international law (“PIL”) provisions on civil liability, in general, and on product liability in particular, are scrutinized, following an approach de lege lata and de lege ferenda.

The concluding remarks integrates the results reached in the analysis and ethical considerations. Both substantive and PIL provisions should be ethically oriented and abide, and ensure, the protection of fundamental rights; private international law shall be an effective instrument for reaching the results pursued by the corresponding substantive provisions. Accordingly, this book will conclude suggesting anew direction of European private international law provisions; as per AI-systems field, it might be time the European legislator accepts connecting factors oriented more towards human rights protection.