Permissioned vs. permissionless blockchains
Not every blockchain is Bitcoin. And not every problem needs a blockchain at all.
A permissionless blockchain, like Bitcoin or Ethereum, lets anyone read it, send transactions and help run it. A permissioned blockchain limits those roles to approved members, as enterprise ledgers such as Hyperledger Fabric and Corda do. Permissioned ledgers are faster and more private but rely on trusting the members. Several high-profile enterprise blockchain projects have been shut down, which is a useful reminder that a shared ordinary database is often the better tool.
The difference
| Compare | Permissionless (public) | Permissioned (private or consortium) |
|---|---|---|
| Who can read it | Anyone | Members, sometimes regulators |
| Who can validate | Anyone with hardware or stake | Approved organizations |
| Identity | Pseudonymous addresses | Known, legally identifiable members |
| Consensus | Proof of work or proof of stake | Usually BFT voting among members |
| Native coin | Yes, to pay fees and reward security | Usually none |
| Speed and privacy | Slower; everything public | Faster; data visible only to those who need it |
| Main trust assumption | Most of the hashing power or stake is honest | The members (and whoever runs the software) are honest |
Enterprise ledgers
- Hyperledger Fabric, an open-source project hosted by the Linux Foundation, lets a group of companies run a shared ledger with private channels between subsets of members.
- Corda, from R3, was designed for regulated financial institutions; data is shared only with the parties to each deal.
- Bank networks such as J.P. Morgan's Kinexys (formerly Onyx) run permissioned ledgers for moving deposits and payments between institutional clients.
There's also a middle path: companies issuing assets on public chains with permissions built into the token. Tokenized money market funds and Treasury funds from large asset managers, for example, run on Ethereum and other public networks but only approved investors can hold them.
The honest limits of "blockchain for everything"
Two well-documented examples:
- ASX CHESS replacement (Australia). The Australian Securities Exchange spent about seven years building a distributed-ledger system to replace its settlement platform, then scrapped it in November 2022 and took a pre-tax write-off of A$245–255 million after an independent review by Accenture raised serious concerns about the system's complexity and readiness.
- TradeLens. Maersk and IBM's shipping-document blockchain was discontinued on November 29, 2022, and taken offline in early 2023. Maersk said the global industry collaboration it needed was never achieved.
Neither failed because "blockchain doesn't work". Both show that the hard part is usually agreement between organizations, data quality and ordinary project delivery, which a ledger doesn't solve by itself.
Do you actually need a blockchain?
Researchers Karl Wüst and Arthur Gervais boiled it down to a few questions (2018). Roughly:
- Do you need to store shared state? If not, you don't need a database at all, let alone a blockchain.
- Are there multiple parties writing to it? If only one, use a normal database.
- Can you use an always-online trusted third party? If yes, let them run the database.
- Are all the writers known? If not, you need a permissionless blockchain.
- Are all known writers trusted? If yes, a shared database is enough. If not, a permissioned blockchain may fit.
Public blockchains are most clearly useful where participants don't know or trust each other and no single company should control the record: open money like Bitcoin, stablecoins and DeFi.
Example: a supply-chain ledger
A food company wants to trace lettuce from farm to store. A blockchain can show that a record was written by a particular farm's key at a particular time and hasn't changed since. It can't show that the lettuce really came from that farm, or that the temperature sensor wasn't faulty. That's the oracle problem again: garbage in, permanently recorded garbage out.
Limitations of each kind
- Permissioned: members can collude to change the record; adding a ledger to a consortium doesn't remove the need to trust its members.
- Permissionless: public data, variable fees, irreversible mistakes and token price risk.
- "Private blockchain" marketing sometimes describes what is really a normal database run by one company.
Frequently asked questions
Is Ethereum permissioned or permissionless?
Permissionless. Anyone can read it, send transactions, run a node, or stake ETH to become a validator.
Do permissioned blockchains have a cryptocurrency?
Usually not. They don't need a coin to reward strangers for security, because the members are known.
Is a permissioned blockchain just a database?
It's a database shared among several organizations, with cryptographic records of who wrote what. Whether that's better than a conventional shared database depends on how much the members trust each other.
Why did the ASX blockchain project fail?
The ASX ended it in November 2022 after an independent review raised serious concerns about the system's complexity and readiness. It took a write-off of A$245–255 million.
Can a company issue assets on a public blockchain?
Yes, and many now do, adding restrictions in the token's smart contract so only approved investors can hold it.
Where to next
Sources
- Blockchain technology overview, NISTIR 8202 (NIST)
- Do you need a blockchain? (Wüst and Gervais, 2018)
- Hyperledger Fabric documentation
- Maersk and IBM to discontinue TradeLens (Maersk, November 29, 2022)
- ASX takes A$250m hit after scrapping DLT-based CHESS replacement (Finextra)
Written for beginners and checked against the sources listed. Last reviewed November 4, 2026. This page is educational and is not financial, tax or legal advice. How we write and check pages.