Robojournalism – A Copyright Study on the Use of Artificial Intelligence in the European News Industry

Abstract

Abstract The copyright protectability of outputs generated by, or with the help of Artificial Intelligence (AI) is a hotly debated question in academia and by many institutions. In practice, sophisticated AI algorithms have become a meaningful assistant in the European news industry for the reporting of sports (Retresco’s collaboration with the German Football Association), weather (textOmatic’s collaboration with FOCUS Online) and finance (the Guardian’s ‘Guarbot’). Furthermore, for the first time in copyright history a court in China assessed the validity of a company’s copyright claim over the articles produced by the corporation’s algorithm. The protection with copyright of this ‘robojournalism’ is no longer just a buzzwordy trend. From a technological perspective, robojournalism currently relies on assistive, generative and distributive technologies. The first two seem to be the most problematic from a copyright perspective as they challenge the well-rooted human authorship requirement. Experts have been able to agree so far that it does not look like AI technology is going to be a disruptive force in the media industry. However, researching the impact of AI in journalism matters a great deal. There are numerous benefits stemming from the use of AI in the newsroom – from expanding news coverage, through faster content production, all the way to leaving journalists more time for the more ‘creative’ and investigative tasks where the algorithm remains weak. This paper addresses, first, the protectability of the outputs of robojournalism under the existing European Union copyright laws. It then goes on to introduce findings related to the practical significance of robojournalism in the European news industry. Here, our focus is on the business, media, and communications studies’ perspectives of automated journalism. Our results demonstrate that the extent to which European journalism relies on assistive and generative technologies to produce written output does not justify, from a copyright perspective, the changing of the current anthropocentric copyright system. These findings have wider implications as AI-generated outputs have prompted many to talk about market failure if copyright (or related rights) protection was to be refused for such works.1