Bug them, use voice recognition technology to translate the speech to text (this is currently available all over the place), use data mining techniques to flag individuals of a certain mentality (maybe sweeten the pot by psychographing your media to cater to certain groups), detect the undesirables. How does the public detect the bugs? Probably can't, who knows what the latest and greatest eavesdropping technology looks like, what kind of frequencies they can operate within, etc. It would take work to create such a system but it's not rocket science.