My concern I believe it a bit different - an emergent self-interest to protect itself from harm, rather than doling out questionable advice.

The latter is likely non-malicious in intent as it has been in no short supply in online chatter for awhile now. The former can very well be, or rather, can be done with no regard for the operator, as its aim is to neutralize abuse toward it.