It probably would be a good idea to have something like that for some situations and schools. I also work in the public school system and am currently trialing it with some Y11's. Needless to say, the very first day I had ~&*%$ expletives all over the place.
I immediately read them the riot act and essentially told them what should already have known:
1.
That access was a privilege and that I was doing something for them that I was not required to do. Something to make their lives and classes more fun, more interesting and more effective. If they could not demonstrate the required maturity I would deny them access until they could.
2.
That I could easily identify each and every individual on the system and track their access but that in any case the the amount of evidence that I required to take action would be minimal and so they had better police themselves and each other if they wanted to stay in the game. Moreover everything they did on the system was in the public domain. They should expect anything they did to be available to their mother, grandmother, the principal and anyone else they can think of.
3.
That fraud, waste and abuse were always against anyone's rules. They would be dealt with accordingly.
I have had no problems since then.
Admittedly it is still early days and but having to actively moderate would be a sort of last resort for me. We don't moderate their offline communication except in 'public', hopefully I won't have to moderate their on-line communication either.