The consistent trend going all the way back to the Payne Fund Studies is that someone somewhere thinks they can show some alarmist correlation between media and behavior.
However, the trend always disappears once you've accounted for shoddy research and disregarded short-term effects (i.e. hormone responses). I could cherry-pick you some recent results; the most popular lately seem to be video games and violence or video games and sexism. It wouldn't be very convincing since you can ostensibly find a variety of papers that say exactly the opposite. I urge you to browse through the google scholar results and draw your own conclusions.
Please do note: I'm arguing a subtle distinction here between subconscious subversion and conscious consideration. After seeing a horrifying documentary about war, I might consciously change my philosophy on death because the source is reliable and the information is believable/consistent. However, seeing silly deaths in a video game can't influence me in the opposite way because the source is dismissible and the conclusion is fantastical.
So, to answer your second question in an unsatisfying way: I think it's clear that people change their own philosophies.