Glossary
Long-Range Attack
A proof-of-stake threat where an attacker rewrites history from an old block using past validator keys, building an alternative chain that can fool newly joining or offline nodes.
A long-range attack targets proof-of-stake chains by rebuilding history from a block far in the past. An attacker who controls validator keys that were active back then — often keys whose stake has since been withdrawn — can sign an entirely alternative chain at little cost, since the old signatures are still cryptographically valid.
The danger is greatest for a node syncing for the first time or returning after a long time offline, which has no recent honest checkpoint to anchor it and might accept the forged chain. Networks defend against this with weak subjectivity checkpoints, key deletion, and time-bound slashing so that distant rewrites become infeasible or detectable.