How does the internet make sure that only entities who own an IP address can rout it to themselves?

Question

When companies want to advertise to the internet that packets meant for a specific IP address should be routed to them, I'm assuming they use BGP, though I only have a vague notion of how BGP works (Kind of like the Dijkstra search algorithm), but what keeps a random Joe from telling the world that packets for 8.8.8.8 should be routed to them, for example?

Answer 1

In theory: BGP filters based on information from routing registries (the IRR), as well as cryptographic verification based on RPKI.

In practice: nothing much.

Good ISPs do filter what prefixes a customer may advertise (though often the filters are built manually), but hijacks, both accidental and malicious, still pop up in the news every so often. There is a Wikipedia article about it.

(Though I would assume sometimes it doesn't work purely because the legitimate paths end up being shorter (and therefore more preferred) than the false ones. I may be wrong.)

And of course, in order to use BGP you need to already own an AS-number and have peering arrangements with some ISP, so that alone raises the bar from "average Fred" to someone who has signed a few contracts – hopefully not under a false name, too.