If "listen to" is something that a person does, does that allow that "speech recognition/transcription" is something a machine can do? If I were servailing en

If "listen to" is something that a person does, does that allow that "speech recognition/transcription" is something a machine can do? If I were servailing en masse, I'd prefer to start with some machine learning filters against all that audio. Is it legally a search if a human didn't do it? What if it's meant to establish "foreignness" percentage? I've noted a lot of "no one's listening to your phone calls," and it seems the sort of "least untruthful" denial a clever government might use.