Causal consistency is a consistency model for distributed systems in which operations linked by cause and effect are guaranteed to appear in the same order on every node, while unrelated operations may appear in different orders on different nodes. For example, a write performed after reading an earlier write will always appear after that earlier write everywhere.
It is weaker than strict serializability or linearizability, which impose a single global order on all operations, but it can be implemented without synchronous agreement among nodes, giving lower latency and easier scaling. It is used in collaborative tools, geographically distributed databases, and large storage systems where a global order is not required.