Of course not, but the whole point of alignment is that an intelligence, human or artificial, understands their ability to do unaligned things but still decides not to.

This would be more comforting if the party in charge of alignment weren't a megacorp trying to maximally extract profits from its workers and customers, but hey, that was also true before AI came along.