Why claims, and why now
AnVerified's shape traces to a $2B verify-once company and its honey-pot problem, a commoditised attestation economy, the World Bank's digital-ID gap, and Malaysia's amended PDPA making data-holding expensive — with one doctrine born at home.
What the record actually says
One market precedent with a warning attached, one proven attestation economy, one global gap, one law change — and the join nobody can buy.
Verify-once identity is a nine-figure business, not a science project
ID.me proved the economics of turning institutional facts — employed, enrolled, veteran, nurse — into reusable, checkout-ready credentials at hundreds-of-millions-of-users scale. It also proved the reliance problem: one company holding everyone's verified identity is exactly the honey pot (and political lightning rod) the model creates. AnVerified takes the credential economy and removes the honey pot from the design.
ID.me · $340M raised, $2B valuation (2025)
Source-attested status already gates the discount economy
Every serious student-discount network resolves 'is this person really enrolled' against registrar data, silently and annually. The attestation pattern is commoditised; what remains unsolved is cross-institution reuse without a central profile — the exact gap an issuer-keeps-the-record, party-receives-a-token protocol fills.
UNiDAYS · Student Beans · SheerID
Most humans are legally identified; barely half are digitally
Roughly 86% of the world's adults hold legal ID but only about half hold a digital one — and legal ID alone solves nothing for the practical question every institution actually asks ('is this person eligible for this thing?'). The market is not identity; it's attestable claims on top of identity. Malaysia, with near-universal MyKad coverage, sits precisely in the gap between documented and machine-verifiable.
World Bank ID4D global dataset
Holding personal data just became expensive — verification tokens didn't
Malaysia's amended PDPA regime raises the cost of every copy of personal data an institution keeps 'just in case': mandatory breach reporting, steeper penalties, real enforcement appetite. An institution that answers 'enrolled? yes' from a signature and stores nothing lives on the far side of that liability. The compliance motion sells the protocol by itself.
PDPA Amendment Act 2024 (breach notification, harsher penalties)
National digital-ID rails exist — the missing layer is claims, not documents
Government digital-identity work proves the appetite and builds the base rails. AnVerified is deliberately not a competitor for the registry seat: it is the relying-party layer that turns institutional facts into scoped, expiring tokens — integrating with whichever national rail exists rather than pretending to replace it.
MyDigital ID initiative (MDEC) · federal digital identity efforts
The join is the thing buyers fear — and the thing the protocol cannot do
Every identity breach headline, every profiling scandal, every 'the app knows too much' story traces back to a universal identifier with a join across contexts. Antrex's own documented doctrine — two ledgers sharing no join key — is this product's spine. It is now stated as mathematics: pairwise pseudonyms make the join non-reconstructable, not merely forbidden.
Design finding: no-join-key doctrine (Masrif two-ledger architecture)
So AnVerified is built this way
The source keeps the record
No central identity vault exists to breach: registrars, HR systems and gatehouses attest and forget. AnVerified signs statements about facts; it never warehouses the facts themselves.
Tokens are pairwise and short-lived
Each relying party sees a different identifier, valid for a scope, expiring with eligibility. Correlation across services is not disallowed by contract — it is impossible by construction.
Integrate with national rails, never duplicate
Where a government identity layer exists, AnVerified consumes it as one attestation path among many. The product is claims and routing, never a parallel registry — politically survivable, practically honest.
Sell the shrink, not the store
The pitch to institutions is data they can stop holding, breaches they can't have, and audits that fit in a signature chain. Minimum storage is the feature; verification is the by-product.
Human-readable and revocable
People see every credential, every scope, every 'who asked' — and revoke in one tap. An identity system without a visible off switch is infrastructure for surveillance with nicer UX.
Assisted paths for the unrecorded
Paper-attested field officers, gatehouse captains and school clerks issue through the same protocol. Digital-first never means digital-only; exclusion is a design bug, not a demographic.
The Malaysian numbers
The person proves who they are once — at the registrar, the ministry, the gatehouse — and every later need is a signature check, not a document upload.
Breach notification and harsher penalties push every Malaysian institution toward the posture that hurts least: hold less, attest more.
The gap between legal and machine-verifiable identity is the addressable market — and in Malaysia it is a claims layer, not a documents layer, that's missing.
Not a policy promise. Pairwise identifiers mean the join key doesn't exist to be sold, leaked or subpoenaed — the architecture is the assurance.
One enrolment week, honestly described
A registrar signs “enrolled” once in September. Through October that student pays staff-group prices at a kedai whose till never sees her name, opens a AnKantin lunch account for her sibling as a verified parent, and passes a library entry gate at campus. In February she takes an internship — the employer gets an expiring “currently enrolled: yes” and nothing else. In June she graduates; every credential quietly dies with the enrolment that spawned it. No vault held her. No service could join her. The only thing that travelled was the signature.
The doctrine this product publishes was born on Masrif’s two ledgers. Read it at the source.
The no-join-key original — Masrif ecosystem→Ready to build something intelligent?
Let's engineer your next product — from idea to launch.