Abstract

Neil Levy argues in a recent JME ‘Current controversy’ paper that responsibility is not an adequate authorship requirement for human researchers, which makes it unjustified to require it from artificial intelligence contributing to research and scientific paper production, although he softens his stance towards the end and accepts that a limited responsibility requirement might after all be reasonable. The main argument provided by Levy against a more extensive responsibility requirement in science is that there are many cases where not all researchers listed as coauthors can assume responsibility for the entire paper or even the central research questions. In this reply, we argue that the more limited responsibility requirement is the ethically reasonable one to ask of all authors, considering the conditions for and value of collaboration, and that this should also have ramifications for the legal regulation of scientific misconduct.