It’s not so simple, unfortunately. The sheer amount of data they have - 212 PB as of December 2021 - makes it practically impossible for most people to mirror. Unless they physically hand over all 745 server nodes to another operator, there’s no way of someone
There are some solutions to this - for example Archive-Team has proposed a method of mirroring the Internet Archive using distributed clients, although this method currently only has a fraction of the total dataset. Still, at this point in time, there’s no real solution to resharing IA’s data in the event they go under