A distributed ledger records data not on one central server but simultaneously across many independent computers, called nodes, each holding a copy that stays in sync with the others through a shared set of rules. Instead of a bank or registrar acting as the single source of truth, the network itself agrees on what counts as valid, so trust comes from cryptography and coordination among participants rather than from one institution's word.
The most familiar implementation is blockchain, which groups transactions into blocks and chains them together in strict chronological order, making past entries effectively immutable. But blockchain is only one design. Ledgers built as a directed acyclic graph (DAG), such as those used by IOTA, skip blocks entirely: new transactions reference and confirm earlier ones directly, letting activity settle in parallel rather than one block at a time. Hashgraph, used by networks like Hedera, takes a related but distinct approach, using "gossip about gossip" and virtual voting to reach agreement quickly without proof-of-work mining.
What unites these variants is decentralization and shared record-keeping, achieved through a consensus process rather than a central administrator. This removes any single point of failure: if some nodes go offline or act maliciously, the rest of the network can still verify the ledger's integrity. Enterprise consortia have also built permissioned distributed ledgers, such as Hyperledger Fabric, where only vetted organizations run nodes, trading some openness for speed and privacy in supply chain, trade finance, and interbank settlement use cases.
Trade-offs remain: coordinating many independent nodes is harder to scale than updating one database, and different DLT architectures balance throughput, security, and decentralization differently.