Automated editorial control Responsibility for news personalisation under European media law

News personalisation allows social and traditional media media to show each individual different information that is ‘relevant’ to them. The technology plays an important role in the digital media environment, as it navigates individuals through the vast amounts of content available online. However, determining what news an individual should see involves nuanced editorial judgment. The public and legal debate have highlighted the dangers, ranging filter bubbles to polarisation, that could result from ignoring the need for such editorial judgment.
This dissertation analyses how editorial responsibility should be safeguarded in the context of news personalisation. It argues that a key challenge to the responsible implementation of news personalisation lies in the way it changes the exercise of editorial control. Rather than an editor deciding what news is on the frontpage, personalisation algorithms’ recommendations are influenced by software engineers, news recipients, business departments, product managers, and/or editors and journalists.
The dissertation uses legal and empirical research to analyse the roles and responsibilities of three central actors: traditional media, platforms, and news users. It concludes law can play an important role by enabling stakeholders to control personalisation in line with editorial values. It can do so by for example ensuring the availability of metrics that allow editors to evaluate personalisation algorithms, or by enabling individuals to understand and influence how personalisation shapes their news diet. At the same time, law must ensure an appropriate allocation of responsibility in the face of fragmenting editorial control, including by moving towards cooperative responsibility for platforms and ensuring editors can control the design of personalisation algorithms.