Why IPFS Alone Doesn't Guarantee Permanence · 1/2
Addressing content isn't the same as hosting it forever
IPFS gives files a durable, verifiable identifier, but it doesn't automatically keep every file stored forever. A node only holds a copy of data it has chosen to keep, and by default, nodes can garbage-collect files they're no longer actively using to free up space. If no node on the network is holding a copy of a particular file, that file effectively becomes unavailable, even though its CID still exists and would work perfectly if someone hosted the content again.
This is a common misunderstanding: uploading something to IPFS once doesn't mean it lives on the network permanently. The protocol handles addressing and retrieval, it doesn't by itself handle the ongoing decision of who keeps paying to store the data.