At least I don't have to worry too much, I never post videos of my surroundings on teh intertubes.

Well, at least there isn’t a microphone in every single device ever ;P

I'd imagine adding random samples of varying frequency and dBi into the audio stream sort of like those AI-spoofing object tracking sweaters may work

Don’t watch the last three minutes of that video then, it gets even more terrifying :))