The version of Laplace's Demon that you're worried about is a part of the universe itself, which means it has to simulate itself, which is as complex as simulating the universe itself, because that's what the demon is doing. It's imaginable that it compresses this (like a quine) but the demon you're worried about also makes decisions that affect the universe, so it has to be able to consider more than one potential future universe, so it can't just assume the copy of the universe inside itself is the same as the one it's simulating outside itself, so it needs to simulate itself simulating multiple universes in which it's simulating...

As far as we know, there is no way to simulate the entire universe that is less complex than the entire universe. Which means that a decision-making Laplace's Demon has to be more complex than the upper bounds on computational complexity.

If we instead assume that it is possible to simulate the universe in less complexity than just having the universe (I don't know how that'd be possible, but it's not formally ruled out by anything as far as I know), then it's possible to fit a more complex simulation inside a less complex one. This means there are infinitely deep hierarchies of simulations, so we must be in a simulation already (as the chances of being at the top are one in infinity). And you'd be instantiated in multiple simulations (infinity is very big), with the chances of you experiencing any of them being proportionate to however popular it is to run that particular simulation, and whichever version of Laplace's Demon you're worried about is probably not a particularly popular one to run.

So in conclusion: either it's physically impossible, or we all live in an infinite number of simulations and the chances you're in the one you're worried about are vanishingly unlikely.