What does blockchain add to digital identity?

Continuation from SSI - Digital Identity Problems.

When I first started working with blockchain identity systems, I was sceptical. I understood the problem with centralised identity, but I was not convinced that putting identity on a blockchain made the user experience or the trust model better.

The useful question is more limited than “can blockchain solve identity?”. It is: which part of the relationship between a person, an issuer and a verifier benefits from a shared, tamper-resistant record? In some cases, that record can help. In many others, it adds cost and complexity without changing the important part of the problem.

What a blockchain can actually provide

A blockchain can help different parties agree on a small amount of shared state without asking one organisation to maintain the only authoritative database. In an identity system, that state might include a public key, a list of recognised issuers or the status of a credential.

That is useful for coordination. It is not the same as proving that a person is who they claim to be. A blockchain can show that a key signed a message or that a record has not been changed. It cannot tell whether the organisation behind the key was honest, whether the original document was correct or whether the person using the key is still the person who received the credential.

Cryptography proves relationships between data and keys. It does not create truth on its own.

This distinction matters because “trust through mathematics” sounds cleaner than the real system. The maths can be reliable while the issuer, wallet, verifier or recovery process is badly designed.

From accounts to credentials

The more interesting change is not that a blockchain stores an identity. It is that a person can hold credentials issued by different organisations and present them to services without creating a new account in each one.

A university could issue a credential about my degree. An employer could issue one about my role. A professional association could issue one about my membership. I could keep those credentials in a wallet and present a specific claim when a service asks for it.

The issuer still matters. The verifier still has to decide whether it accepts that issuer. The wallet still has to protect the credential and offer a way to recover access. Blockchain may support parts of this coordination, but it is not the identity itself.

Selective disclosure is the user benefit

The clearest benefit is reducing the amount of information I have to reveal.

If I need to prove that I am over 18, I should not automatically have to send my full driving licence. If I need to prove that I have a degree, the employer does not need every detail printed on the diploma. Selective disclosure can let me present only the relevant claim. Some credential systems use zero-knowledge proofs to achieve this; others use different cryptographic formats.

The technique matters, but the user-facing result matters more: the service receives the smallest useful piece of evidence instead of a complete document.

There is an important limit. Selective disclosure does not stop a verifier from storing the claim after I present it, asking for additional information or combining it with other data. Privacy depends on the protocol, the interface and the behaviour of the organisation receiving the proof.

Revocation is not a kill switch

One of the claims around blockchain identity is that users can revoke access to their data. This needs to be stated carefully.

A credential can be revoked. A verifier can stop accepting a credential. A permission can be changed. None of that deletes information that has already been copied, recorded or inferred by another service.

If I give a company my address, I cannot use a transaction to make the address disappear from its database. Revocation can change whether a credential is valid in the future; it cannot undo a disclosure that already happened.

This is also where permanent public records become uncomfortable. Putting identifiers or identity events on a permanent ledger may help with auditability, but it can conflict with privacy, correction and deletion requirements. A system that is difficult to alter is not automatically a system that is good for people.

Where the possibilities are real

There are areas where these tools could be useful, provided the surrounding governance works.

In finance, a user could present a credential about income, residency or credit history without handing every service a complete financial file. That could reduce repeated checks, but it would not eliminate the need for underwriting, regulation or judgement. On-chain activity is not the same as a reliable credit history, and a wallet address is not a complete picture of a person.

In supply chains, credentials can link a product to an issuer, a batch, a certification or an event in its history. That may make provenance easier to check. It does not prove that the information entered at the beginning was true. The system still depends on inspections, incentives and the people or machines that supply the data.

In healthcare, portable credentials could help patients prove access rights, professional qualifications or consent across institutions. That is different from putting medical records on a public blockchain, which would create serious privacy and governance problems. The useful layer is usually the exchange of permissions and claims, not the publication of the medical data itself.

Voting is a much harder case. A credential could help establish eligibility, but it does not by itself solve coercion, ballot secrecy, device security, unequal access or the need for a trustworthy counting process. “Blockchain voting” is not a shortcut around those problems.

New business models, with a warning

This architecture could create services around credential issuance, verification, wallet recovery and privacy-preserving analytics. Some organisations may provide infrastructure without storing the underlying identity data.

That is a possible business model, not a guaranteed outcome. A verification provider can become a new gatekeeper. A wallet provider can become a platform. A service that promises privacy can still collect metadata about where and when credentials are used.

Even the idea of monetising personal data needs caution. Giving a user a payment for access does not necessarily give them meaningful control, especially when the service is essential or the terms are difficult to negotiate. “The user owns the data” is not enough if the user has no practical alternative to sharing it.

The problems that remain

The user experience is still the hardest part. A person should not have to understand public keys, signatures, credential schemas or revocation registries to prove a simple claim.

Key management creates a different kind of responsibility. If I lose access to a wallet, what exactly do I lose? Who can help me recover it? Can a recovery provider restore access without becoming powerful enough to impersonate me? A system that gives me control but no safe recovery path is not user-centred; it is just transferring operational risk to me.

There are also problems with linkability. If I use the same identifier or wallet everywhere, different services may be able to connect my activity. Privacy-preserving credentials need to limit that correlation, not only hide the contents of one document.

Adoption is another constraint. Users will not carry credentials that no service accepts. Services will not integrate credentials that too few users carry. Institutions also have reasons to keep their current systems: they understand them, they control them and their business models may depend on that control.

Standards can improve interoperability, but they do not settle questions about liability, issuer governance, accessibility, regulation or support when something goes wrong. Nor do they make different systems understand the same claim automatically.

A more modest role for blockchain

I do not think blockchain identity will replace documents, banks, governments or platform accounts overnight. I am not even sure that it should replace all of them.

The more realistic model is hybrid. Existing institutions can continue to issue credentials. Users can hold and present them through wallets. Shared registries can help verifiers check issuers, keys and status. Blockchain can be used where a shared record adds real value, and left out where a normal database is simpler, cheaper and easier to correct.

For me, the test is practical. Does the system reduce the amount of information I have to disclose? Can I understand what I am presenting? Can I recover access? Can I use the credential in more than one context without being tracked everywhere? Can I challenge an issuer or verifier when they make a mistake?

If the answer is no, adding blockchain does not change the underlying relationship. It only gives the existing problem a more fashionable vocabulary.

---

Next in the series: Which use cases and products are being built?