I have been coming round to the once-crazy-sounding idea that the only way to protect consumers from unwittingly bugging themselves and others around them is to require all devices to state a prominent warning about any components they include that can be used as sensors or communication devices, and to include at least a hardware-driven indicator of when any sensor or communication channel is active and preferably a hardware switch to force it off.
Whether any government would ever support such a move is a different question, of course. I suspect not in the current climate, because they'd fear losing their own intelligence capabilities against targets they wanted to single out legitimately. However, I also get the feeling that the tide is finally turning against the "mass surveillance/database state" in the general public consciousness and not just for geeks and civil liberties campaigners. Probably not at the next major electoral cycles but maybe in the ones afterwards, I suspect preserving personal privacy will be a political issue with some real weight, and this kind of issue will be part of that debate.
Even the accelerometer in a phone can be used to record keystrokes from a keyboard. http://www.i-programmer.info/news/105-artificial-intelligenc...
Even WiFi can potentially be hacked to map/pinpoint people in a room/building.
Must be a fun time to be a spy with everyone putting numerous bugs, sensors, and cameras in their work and home willingly.