To me, the idea of an algorithm which detects when a driver is at increased risk (e.g. drowsy, or not wearing a seatbelt, or maybe using a phone inappropriately) seems like a fundamentally good move. And makes basic sense in any vehicle still piloted by a human.
The issue is that (of course) people being watched by a camera opens the potential of misuse, and so (as ever) people jump to worse-case scenarios and start using emotionally-loaded words like ‘dystopian’.
Of course, we don’t know for sure if it will be misused, and/or dystopian, until it’s actually implemented and we have observed results... but people do like to get upset and sensationalise just about anything that’s a new idea, these days.