Makes sense to me. Unfortunately, it all breaks down if you have a flight crew that’s doing the nefarious activity.
This true, but short of us developing an intelligence on the order of the fictional SkyNet, I don’t see a workaround for having to trust human beings in the cockpit.
I don’t mean to say that we don’t have rudimentary AI that could probably completely handle a plane in just about any condition, but the truth is, we are going to have human pilots for the foreseeable future, until the machine pilots can B.S. with passengers & flirt with flight attendants.
Humans like humans being in control, because we can understand and interact with them, even if the humans may be more fallible or less trustworthy! I think this boils down, at its base, to the fact that most believe the human pilot wants to live as badly as they do, and they don’t understand programming concepts or how the machine does or doesn’t perceive the human condition.