Artificial intelligence and civil liability: EURA recommendations

The European Centre of Excellence on the Regulation of Robotics & AI of Sant’Anna in Pisa has published an interesting document, under request of the Legal Affairs Department of the European Commission.

It is now clear that AI software has entered our lives in all areas, including health care. As it is now clear that the undoubted contribution of these technologies is accompanied by many doubts and perplexities about their invasiveness, the possibility of discrimination, the real use that will be made of our data, the harmful effects and related damage that may result.

All those issues have found a first attempt of analysis in the consultation of the European Commission on February 19, 2020 through the publication of the “White Paper on Artificial Intelligence“, where the reflection takes its cue and develops around a central node: the use of AI (as any new technology) involves both risks and opportunities, it is therefore necessary to reason in order to define a legal framework capable of facing the new challenges.

As regards in particular the regulation of civil liability, there is a very recent document commissioned by the Legal Affairs Department of the European Commission to the EURA “European Centre of Excellence on the Regulation of Robotics & AI” of Sant’Anna in Pisa entitled “Artificial Intelligence and Civil Liability” published on July 14.

The document is really very interesting and certainly deserves a mention. After acknowledging that the regulation on civil liability – also of IA – is contained at EU level in Dir 85/374/EEC, the document goes so far as to suggest possible solutions for the specific sector of IA.

More precisely, in addition to suggesting in general an update of the contents of the Directive (also transforming it into an EU regulation to make it legally stronger), the document also suggests that the general framework should be flanked by specific rules for technologies that may present risks and potential considered relevant.

But the most interesting part of the document is undoubtedly the one that suggests to review the logic of the legislative approach to civil liability in the IA sector.

In fact, it is recommended to move from a liability system based on fault to a logic based on risk management: this is in line with the new legal architectures not only of product (think of the new EU Reg. 2027/745 on medical devices) but also of process (in this case the reference to GDPR is a must).

In essence, attribute the liability (and therefore the burden of paying damages) to those who actually in the process of design, creation, management and use is able to “affect” the risk and “measure” the risk/benefit ratio.

A logic that is therefore very different from the current one of our Civil Code, which is much closer to that – to which we are already accustomed – of the health risk approach and risk management.

Silvia Stefanelli

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s