Tekoäly ja vahingonkorvausvastuu media- ja viestintäalalla: teoreettisia lähtökohtia ja valikoituja havaintoja
Avainsanat:
vahingonkorvausoikeus, sopimusoikeus, tekoäly, viestintä ja mediaAbstrakti
Lakimies-aikakauskirjan parhaan kirjoituksen palkinto 2019
Artificial intelligence and tort liability in the media and communications industry: Theoretical considerations and selected examples
The article examines artificial intelligence (AI) and tort liability in the media and communications industry. First, the authors discuss overarching theoretical questions in terms of regulation of AI-related harms, and then proceeds to address two examples, namely automated journalism, or publications produced by “robot editors”, and tackling illegal online content. They look at these areas from the standpoint of Finnish and European law. The article aims to highlight some interesting AI-related liability questions and intriguing details pertaining to the possible significance of the utilisation of AI for tort liability considerations. As background, the authors point out that (particular) regulation of AI-related harms is a topic for heated discussion both in Europe and globally, and different kinds of tort liability solutions, including solutions involving strict liability, have been put forward as preliminary ideas for addressing AI-related harms in the future.
The methods used in the article are mainly legal doctrinal and theoretical legal doctrinal. The authors conclude, in particular, that there are not yet specific Finnish rules that would target liability in AI-related situations such as those they use as examples. If an AI-related harm occurs in an example situation, the applicable law is mainly Finnish tort law. If AI has a significant role in causing a particular harm, this may render applying some tort law rules challenging, at least in theory, but in practice it appears that at least in the example situations resolving AI-related tort cases would, in the end, be rather straightforward. Nevertheless, arguments that seek to support avoiding liability can be based on e.g. the unpredictability of AI. The fact that the example situations are governed merely by general tort law is not necessarily a problem. In the future, the regulation of obligations in utilising AI could be a wise way of (indirectly) clarifying also details of private liability.
