Imagine you are in a tournament no rebuys allowed, you are on the big blind with AA and someone on early position pushes all in, should you take a chance on variance against a probably (AA, AK, KK, QQ) or should you slowly grow your stack?
Having 200 big blinds does not mean that you're twice as likely to win the tournament over someone who has 100 big blinds.
However, your chances of cashing do increase by a good amount and result in you being able to put more pressure on your opponents and at the same time you'll be less likely to becoming short stacked at the same time your opponents will, meaning you won't have to take high risk spots as often as they will.
Folding AA would be leaving a huge amount of equity. You have ~80% equity. If you say no to it you're probably playing a tournament you can't afford. Go all in and go play the next tournament if you lose.
From an ICM standpoint you're not even allowed to fold AA on the bubble (one exception: in satellites where you are near the bubble and guaranteed to make the money), let alone here. Only exception would be if you're guaranteed to cash anyway and a min-cash would be live changing money to you, which is unlikely.
To get an exact answer you can go into solvers such as ICMizer or Holdem Resources Calculator and see what hands you can call all in for and in which situations, such as this one.
I understand the argument of surviving, but I only think it's valid to this extreme if you're not trying to win the maximum in the long term.
You call. The previosu answer is good enough, but there is some more reasoning to it. If you fold, everyone can beat you by just shoving any-two-cards against you (if you have aces, there is no big difference between them having JJ or 76s)
Also, even if you are assuming some risk, most situations assume a much higher risk. You just have to accept that even the best player does not win every time