When you see traffic between 40 yo man and 12 yo girl which don't have any common social connections and the messages are initiated by the man, you don't have to crack e2e to suspect dickpicks.
So you want the platform to be creepier and investigate connections more intensely? And you want to intercede on an arbitrary method you just made up, without examining all traffic first?
I seem to recall someone taking pictures of their baby, naked, because it was sick, and emailing them to the doctor -- and having their Apple account terminated. Terminated, with the father being labeled a pedophile, and the police contacted (all automatically).
Everyone was quite upset. Everyone felt it was too intrusive.
Frankly, communication platforms have no business trying to police anything at all. I wouldn't want the phone company recording all my conversations, hunting for trigger words, and then contacting the police or cutting off my phone if I sad "bad word".
Yet somehow it's OK to have this level of intrusion because.. um "computers".
The state has no business listening in on private citizen's communication.
Corporations have no business doing so.
To protect the 12 year girl, something called "her parents" need to pay attention and watch what she does. That's their job. They're her guardian.
Some random corporation has no business in that. Some random corporation has no business being an 'algorithmic parent', an automated machine with no appeal.
Here's something I'd support -- a way for parents to prevent children from registering for accounts, and, to be able to examine children's accounts.
But... then we get into ID verification. Of course, surely you support ID verification for platforms, because if you support platforms knowing the age of people (40 and 12, you listed), then you therefore must support a way to verify those ages.
> which don't have any common social connections
How would you actually know this? Facebook is a surveillance company, but they are not omniscient.