Blindly accepting AI outputs, setting poor AI controls, failing to update processes when software changes: many AI problems unfortunately begin with the human, but it is also humans that can prevent them.
Humans: the main cause of, and solution to, AI mishaps