That the AI can always get out of the box is a very strong claim. But it's not necessary. The only claim necessary is that the AI can convince some humans to

That the AI can always get out of the box is a very strong claim. But it's not necessary. The only claim necessary is that the AI can convince some humans to let it out of the box, and we cannot identify a priori which humans will and will not let it out, thus we cannot guarantee we'll keep it in the box. That's a much weaker claim, but proves the same general point and is much easier to argue.