While they correctly note that correlation does not imply causation (for instance, it is likely there is a third factor at play, perhaps a psychological one), there actually is a study which contradicts "criminals using more porn": https://link.springer.com/article/10.1007/BF01540933