@maggiejk @johncarlosbaez
My recollection is that the recent research seems to go opposite direction with time. Most systematic reviews, when adjusting for methodology, don't find anything like that.
Note: I'm talking about research into adult pornography and sexual violence in general, because it's studied most. I'm not very familiar with pedophilia.
Again, correlation should not be surprising, causal link is harder to come by.
Review Manuscript
Pornography and Sexual Aggression:
Can Meta-Analysis Find a Link?
https://doi.org/10.1177/1524838020942754
... Nonetheless, evidence did not
suggest that nonviolent pornography was associated with sexual aggression. Evidence was particularly weak for longitudinal
studies, suggesting an absence of long-term effects. Violent pornography was weakly correlated with sexual aggression, although
the current evidence was unable to distinguish between a selection effect as compared to a socialization effect. Studies that
employed more best practices tended to provide less evidence for relationships whereas studies with citation bias, an indication of
researcher expectancy effects, tended to have higher effect sizes. Population studies suggested that increased availability of
pornography is associated with reduced sexual aggression at the population level. ..