Midnight and selective disclosure
The missing piece is not another public record
Community finance needs accountability. It also involves information that should not become public: identity documents, household circumstances and financial histories. Our technical question is whether people can verify the facts a program needs without exposing everything behind them.
Midnight is a blockchain platform we are researching for that purpose. Its Kachina model connects public contract state with private state kept locally, using zero-knowledge proofs to validate defined changes.[14] This is a possible technical foundation for Kapitala, not an announcement of a partnership or deployed product.
What “fourth-generation” means here
Midnight describes itself as a fourth-generation blockchain. Its own account of this progression moves from digital money, to programmable contracts, to scale and interoperability, then to identity and privacy.[22][24] We use the term as Midnight's framing, not as a universal technical standard, certification or proof that one network is best.
The useful contribution is programmable privacy: deciding what must be verifiable, what should stay private, and what may be disclosed for a particular purpose. Midnight calls its combination of privacy and practical utility “rational privacy.”[23] That is more relevant to a community-finance project than the generation number itself.
A shared ledger can help different participants check agreed rules. Making the underlying personal records public is not a necessary goal. Conversely, keeping everything secret would make legitimate oversight difficult. Selective disclosure is a way to design a more useful middle ground.[15]
Selective disclosure and zero-knowledge proofs
These are related, but not identical:
- Selective disclosure means revealing only specified information for a defined purpose, rather than an entire record.
- A zero-knowledge proof can establish a particular statement about private data without revealing the underlying data. For example, a proof could establish that an attribute meets a threshold without disclosing its exact value.[16]
- Explicit disclosure in Compact, Midnight's programming language, requires a developer to mark potentially private information before exposing it through the public ledger, an exported circuit or another contract. The decision still belongs to the application and its design.[15]
The distinction matters: revealing a chosen field is a disclosure; proving a condition about it can avoid revealing that field. Neither automatically creates informed consent, secure devices or fair decisions.
What Kapitala could do with it
These are proposed designs, not operating features or completed experiments.
| Community-finance need | Possible proof or disclosure | What still needs human and technical work |
|---|---|---|
| Establish membership | Prove possession of a current credential from an accepted cooperative without publishing its full membership register | A trusted issuer, identity checks, expiry, revocation and safe recovery |
| Check a program rule | Prove that a calculation used the agreed rule and valid private inputs, without publishing each member's circumstances | Correct rules, trustworthy input records, complete checks and a way to correct mistakes |
| Provide scoped oversight | Disclose the records an authorized reviewer is entitled to inspect while withholding unrelated information | A lawful basis, recipient verification, access controls and accountability for subsequent copies |
| Carry a useful record between programs | Present a limited, consent-based credential about relevant history rather than surrendering an entire financial profile | Fairness testing, correction, portability, issuer acceptance and protection against coercion |
A valid credential or proof does not establish creditworthiness, prove affordability or guarantee approval. A repayment record is not a complete picture of someone's needs or ability to repay. Any future reputation signal must remain exploratory, context-specific and contestable.
Less dependence on promises, not the end of trust
Midnight's developer documentation warns that ordinary ledger reads, writes and operation arguments are public, with a specific exception for its Merkle-tree data types. Applications must deliberately choose what they expose, including information that can be linked across interactions.[27] Merely using the network is not a privacy guarantee.
“Trustless” can describe a narrow property of a cryptographic protocol. It is not an honest description of a whole financial service.
A proof can show that a computation followed a defined rule. It cannot make a false source record true, make an unfair rule fair or guarantee that money was handled properly outside that proof. People still depend on issuers, operators, software, devices, governance and the legal system. Privacy also depends on what an application deliberately discloses and what can be inferred from timing or repeated interactions.
The practical aim is less unnecessary trust and less unnecessary disclosure, with responsibility still clearly assigned.
Why a blockchain rather than an ordinary database?
We should earn that choice, not assume it. A conventional database with role-based access, signed records and independent review may be simpler for a small cooperative pilot. A shared cryptographic system becomes more interesting when separate organizations need to verify the same rules or accept limited credentials without handing each other complete private records.
Our comparison must cover total cost, recovery, usability, security, independent verification and the information exposed. If a simpler system serves members better, the pilot should use it. Blockchain is not a requirement for community benefit.
Making the technology usable
Midnight's wallet documentation describes DUST sponsorship, in which a separate party pays transaction fees.[28] This could be worth exploring so that a future member would not need to manage each network fee personally. It does not make operation free: someone must provide the resource, and funding, security and sustainable costs remain design questions. No Kapitala fee-sponsorship flow or token purchase requirement is announced here.
Device performance is equally important. The wallet documentation supports browser-based WebAssembly proving but describes it as significantly slower than a dedicated proof server.[28] The proof-server guide explicitly warns that information sent to that server includes private data.[30] Moving the computation to an outside service is therefore not a privacy-neutral shortcut. We need real device measurements and a custody model before promising that a low-cost smartphone can do everything privately and conveniently.
Midnight also documents a specific connection from Cardano NIGHT to DUST generation.[29] That is not evidence of a general-purpose bridge for Kapitala, interoperability with every chain or portable financial records already working between cooperatives.
What is built today
Kapitala is not integrated with Midnight. Its blockchain package is a scaffold; current rule-engine and ledger work uses synthetic test data. A local hash-linked record is not a Midnight transaction or a zero-knowledge proof. No live member credentials, private underwriting or blockchain fund custody are offered by this website.
The small Siargao pilot is intended to test the minimum useful service first. Research into Midnight can run alongside it without turning every future technology into a prerequisite for learning.
Before adoption we need a synthetic prototype, a disclosure and threat model, measured performance on low-end phones, a safe assisted-access route, credential and key recovery, independent security checks, and a legal review of the actual proposed service. Proof generation on a private local device and proof generation entrusted to a remote service have different privacy implications; neither is assumed solved.
Continue with the research agenda, connected communities, or transparency and privacy.
Sources
[14] https://docs.midnight.network/concepts/kachina — Kachina | Midnight Docs [15] https://docs.midnight.network/compact/reference/explicit-disclosure — Explicit disclosure | Midnight Docs [16] https://docs.midnight.network/concepts/zero-knowledge-proofs — Zero-knowledge proofs | Midnight Docs [22] https://midnight.network/about — Midnight | About [23] https://midnight.network/blog/midnight-at-consensus-2025-building-the-future-of-rational-privacy — Midnight Blog | Consensus 2025: Building the future of rational privacy [24] https://midnight.network/blog/midnight-network-is-live — Midnight blog | Midnight network is live [27] https://docs.midnight.network/concepts/how-midnight-works/keeping-data-private — Private data | Midnight Docs [28] https://docs.midnight.network/sdks/official/wallet-developer-guide — Midnight wallet SDK | Midnight Docs [29] https://docs.midnight.network/concepts/dust-architecture — DUST architecture | Midnight Docs [30] https://docs.midnight.network/guides/run-proof-server — Proof server | Midnight Docs