Both can be true – roads can be more lethal on a per-mile-traveled (or per-minute traveled) basis, but with many fewer miles/minutes traveled overall, total cumulative fatalities are down. That seems to be the case in these articles.
This sounds like it's in line with the theory of risk homeostasis or at least the Peltzmann effect.[0] The theory states that as people feel safer they will take riskier actions. In this case the roads are emptier so people feel that they can drive faster, which leads to more accidents. Note that the theory does not have much support, but there seems to be some kind of an effect like this.
