Goatse has been online for thirty years and I’ve never seen anybody say “I would definitely have never tried that if nobody showed me that website”
Are you asking for evidence that humans tend to emulate what they see other humans do?
Are you asking the more direct classic question of if there's evidence that violent media correlates with violent acts?
You want evidence that rises to the level of establishing "causality" in consideration of a natural experiment that is being run across all of humanity simultaneously?
What populations are protected from violent media?
How would you even disentangle all the countless confounding factors?
The arguments here are well-worn by the industries that peddle in these types of media, with obvious incentives, and obvious incentives on our side as consumers to not be restricted from consuming whatever we enjoy.
This is why I spent so much time referencing all the other ways in which humanity tends to emulate the behavior of other humans or be influenced by advertising/media, as it seems unlikely that these tendencies would suddenly cease around the sole category of "violent media".