Minimal Disclosure

Minimal disclosure is a design rule for identity systems, not a piece of cryptography: reveal the least amount of identifying information possible, and only what's actually needed for the specific interaction happening right now, never everything a system happens to know, "just in case it's useful." A driver's license that shows your full birthdate, address, and photo to prove you're over 21 at a bar violates minimal disclosure even if nothing about the transaction was insecure: it discloses far more than the bar's actual question ("are you over 21?") required.

This term is easy to confuse with two neighboring terms in this Lexicon, and untangling that confusion is the actual point of this entry:

• Minimal Disclosure (this entry) is a design principle for identity systems: a "why," articulated by an identity architect thinking about how systems should be built.
• Selective Disclosure (a separate Lexicon entry) is the cryptographic mechanism, the "how": that makes minimal disclosure technically possible for verifiable credentials, via SD-JWT or BBS+.
• Data Minimisation (a separate Lexicon entry) is a legal obligation under data-protection law (GDPR Article 5(1)(c)), a "must," imposed on organizations processing personal data, with a different lineage (data-protection regulation, not identity-system architecture) even though it points in the same direction.

All three describe roughly the same instinct, take less than you could, from three different angles: a design law, an engineering technique, and a legal duty.

Who actually wrote the principle

Kim Cameron, at the time Microsoft's Identity Architect, published "The Laws of Identity" in 2005, seven principles for how digital identity systems should be built, written in the aftermath of widespread industry frustration with centralized, security-incident-prone identity schemes, including Microsoft's own earlier Passport service. Law 2 of the seven is commonly summarized as "Minimal Disclosure for a Constrained Use": the idea that the most protective identity system discloses the least identifying information, limited to what a given use actually requires. Solidus did not write this essay and did not invent the principle it names.

How Solidus implements it, technically

The technical mechanism is documented in full under Selective Disclosure, BBS+ signatures and SD-JWT VC, both borrowed cryptography, not invented here, and this entry deliberately doesn't repeat that content. What's worth restating plainly: Solidus's live testnet issuer speaks SD-JWT VC (confirmed via its OID4VCI metadata), and a separate BBS+ path exists on the same testnet but is unaudited, with an external audit targeted for H2 2026 via NLnet NGI Zero. There is no mainnet.

Check it yourself

The two proofs are identical to the ones documented under Selective Disclosure: curl -s https://capture-api.solidus.network/.well-known/openid-credential-issuer confirms the live SD-JWT path; npm i @solidus-network/auth @solidus-network/bbs plus the bundled demo/unlinkability-demo.mjs demonstrates the BBS+ path, with its own printed output stating the audit-pending boundary.