Sign (@Sign): Identity is Governance: Why Verifiable Credentials are About Power

Identity is Governance: Why Verifiable Credentials are About Power Author: Harry Song, Researcher @Sign A modern identity system is a power problem, not a database problem. Imagine walking into any public office. Or a bank. Or perhaps, a clinic. The receptionist asks who you are. You hand over a card. They copy it. They type it. They call someone. They stamp a form. They keep a record. They might then divert you to someone else, where you repeat the whole process. You leave. That little routine looks harmless. It’s not. It tells you what you can see, what you can prove, and what you have to reveal about yourself to do it. It decides who can ask (who’s the requester?). It chooses who can store (who gets to have your data?). It chooses who gets to check (who’s the authenticator?). It decides who can say no to you and how you can appeal. Identity is the permission system of society, and permission systems are always about power. The real question isn't "Who are you?" It is: Who gets to decide? In the analog world, we built guardrails out of friction. A paper file takes a long time. You have to do it by hand with a stamp. A clerk can only do so much work. A cabinet doesn't accidentally show up in five different agencies. Going digital removes that friction. It scales everything, even mistakes, abuse, and even surveillance. So, when governments modernize identity, they aren’t just buying software. They decide the shape of authority for decades. It’s worth saying this plainly: If every service depends on one central gate, then the operator of that gate becomes a super-ministry. If every verification requires a database call, then every verification becomes a log entry somewhere. If every agency copies the same sensitive data, then every agency becomes a breach surface. Those are results of governance, not of technology. Why Digital ID Debates Often Go Nowhere Most identity debates get stuck in a false dichotomy. One side wants maximum efficiency. One system, one number, one truth. It feels clean, and it feels modern. On the other end, that side wants maximum privacy. No tracking, no central visibility, and no single point of failure. This system feels safe and principled. Both sides are right, yet both sides are drearily missing something. A sovereign system has to do two things at once: Deliver state capability. Services must work at national scale, eligibility enforced, fraud reduced, and exceptions handled without chaos. Deliver citizen transparency. People must see what is requested. They must share only what is needed and have a way to contest, correct, and recover. If you design only for capability, you get a honeypot, whereas if you design only for privacy, you get a pilot that never becomes infrastructure. The right target here is controllable privacy that is 1) private to the public by default; 2) auditable to lawful authority when it must be; and 3) operable under real national conditions. That is not a slogan. It is a design requirement. Verifiable Credentials are not magic. They are a new policy primitive. A Verifiable Credential (VC) is simple in spirit. It is a signed statement from an authority. Think of a stamp that can be checked anywhere. It’s basically a credential that says, “This ministry attests that this person holds a valid license,” and that statement can be verified by anyone who is authorized to verify it. Yes, it uses cryptography. But no, you do not need to care about the math. You only need to care about what it changes. 1) It separates proof from databases Today, most verification works like this. First, the verifier collects raw data. Secondly, they store raw data. Then, they call a system of record. And lastly, they keep a log. Here, compliance becomes a paperwork industry (and who likes paperwork?). And this is how personal data spreads, how breaches become national events. With VC s, you can flip the flow: The issuer issues a signed claim. The holder stores it. The holder presents what is needed. The verifier verifies authenticity and status. In plain terms, you can verify a fact without copying the whole file. That one shift is huge for governance. 2) It makes minimal disclosure practical Policy people talk about data minimization. However, in reality, product teams often cannot implement it, and legacy systems do not make it easy. VCs can. A verifier can ask for “over 18” instead of a full birthdate. A service can ask for “resident” instead of a full address history. And a bank can ask for the fields it is legally required to collect, not the fields it happens to want. Minimal disclosure turns privacy from a promise into a user experience, and user experience is where policy becomes real. 3) It creates a cleaner line between legitimacy and surveillance In a healthy system, verifiers should be able to confirm that i) the credential is authentic, ii) the issuer was authorized, iii) the credential is not revoked, iv) the credential matches the request, etc. without turning every check into centralized tracking. This is the heart of controllable privacy, where verification should be cheap, and surveillance should be hard. In all, that is the point. Power Lives in the Trust Fabric If identity is governance, then governance must show up as explicit mechanisms, not informal habits. A VC system forces you to name the real moving parts. 1) Who can issue what A credential is only as legitimate as its issuer. So the state must define i) trusted issuer lists: which organizations are allowed to issue which kinds of claims; ii) delegation: who can authorize sub-issuers (for example, a ministry authorizes universities, a regulator authorizes licensed banks), and iii) key management: how issuer keys rotate, and how history remains verifiable when they do. This is not optional bureaucracy. It is the backbone of trust. 2) What a credential is allowed to mean A credential schema is a policy artifact. It decides which fields exist, which fields are sensitive, what evidence is required to issue it, how long it is valid, and what it means to revoke it. If you do not govern schemas, you do not govern meaning. If you do not govern meaning, you do not govern outcomes. 3) Who can request what Not every verifier is equal. A bar checking age is not a bank onboarding an account. A hotel is not a tax authority, and a benefits office is not an e-commerce app. So you need tiers. You need a way to say who can request which attributes, under what legal basis, with what retention rules, and with what audit obligations. With these cleanly defined, you avoid becoming a monolith without becoming a free-for-all. Audit Is Where Sovereignty Becomes Credible Many systems talk about privacy. Fewer talk about oversight. Governments do not only need to serve citizens. They need to prove they served them correctly. They need to investigate abuse, resolve disputes, and answer parliament, courts, inspectors, and the public. So, a sovereign identity system must produce evidence. Not raw personal data, but real, hard Evidence. Governments need evidence that can answer: who authorized an issuer, who approved a schema change, who revoked a credential and why, who verified what under what authority, and when it happened. The best time to design that evidence is day one. The worst time is after the first scandal. A Concrete Example: Renting an Apartment Let’s keep it simple. A citizen wants to rent an apartment. The landlord needs i) proof of identity, ii) proof of income, iii) proof that the applicant is not blacklisted or under sanction, and iv) sometimes, proof of residency status. Old world: The applicant sends scans of ID. Sends bank statements. Sends an employment letter. The landlord stores copies. The property manager stores copies. The documents sit in inboxes and shared drives. Sensitive data spreads everywhere. In this system, no one is fully sure what is authentic, and everyone keeps more than they actually need. New world with SIGN’s VCs: The government issues a signed identity credential. An employer issues a signed income credential. A regulator issues a sanction-clear credential. The applicant holds them in a wallet. The landlord requests only what is required: proof of identity, proof of sufficient income, proof of legal residency. The applicant consents and presents. The landlord verifies authenticity and revocation status without calling three databases or storing full financial records. No unnecessary bank statements. No photocopied passports in email threads. And no uncontrolled data duplication. The landlord gets confidence. The tenant keeps control. The system remains auditable if there is a dispute. That is what modern should mean. What Decisions a Country Must Make If you are a policy owner, a regulator, or a program lead, here are the decisions that actually matter. Define the trust authority. Who operates the national trust layer? What are their emergency powers? What is the appeal path? Define issuer governance. Who can issue which credentials? How does delegation work? How do you remove an issuer safely? Define verifier governance. Who can request what? How do tiers work? What evidence must verifiers keep, and what must they never store? Define schema governance. Who owns base schemas? Who owns sector schemas? How do versions roll out without breaking services? Define revocation posture. How fast must status be updated? What works offline? What requires an online check? Define recovery. What happens when a citizen loses a phone? What happens when an issuer rotates keys? What happens in compromise? If those questions feel like governance, good. They are. The Closing Truth Identity infrastructure is where power hides in plain sight. Verifiable Credentials do not remove power. Instead, they reshape it by letting states deliver services without building a single database that everyone must trust forever. They let citizens prove facts without handing over their entire lives. They also allow oversight toexist without turning society into a permanent log file. Build it as infrastructure. Treat it as governance. And make privacy controllable; make audit real. That is how a sovereign system lasts. A note on SIGN SIGN works on the layer beneath the apps. We build the trust fabric of who can issue, who can ask, what it means, what gets recorded, and what never should. We care about the boring parts that decide whether a system survives reality: recovery, revocation, offline verification, and audit. We think good identity should feel like good infrastructure. It should be dependable, minimal, and easy to verify. The proofs should travel, the payloads should not. If we do this right, citizens get clarity and control, institutions get accountability, and governments get systems that scale without becoming a honeypot.