Can filtering protect children from harmful material in chatrooms, blogs, and social networks?

I don’t think there is any explicit filtering technology available that can completely avoid harmful material from chartroom, social networking and blogs. There should be manual and AI based monitoring available that checks for content periodically and raise alarms when it detect something bad going on. All such sites should be personalized services so that users are supposed to register and offenders can be easily thrown out of the site.

I wish technology advance to the level that we are able to use retinal detection to determine age so that whoever uses the computer to view a particular website can be verified of age.