The answer is not necessarily “everything.” Midnight’s FAQ states that private state is encrypted and stored locally. A seed phrase can recover unshielded addresses and public history, but it does not by itself recover unspent shielded notes or private contract data. If that local state is lost, the associated assets may become inaccessible.
That is a sharper constraint than the familiar account-wallet model. On an account chain, a wallet restore often means deriving an address from a seed and asking a node for the current balance. On a shielded system, the chain can show that commitments exist without revealing which encrypted commitment belongs to the restoring user. Recovery becomes an indexed private search problem.
Midnight’s launch announcement says the network went live in March 2026 with state continuity and permanence intended for programmable, end-to-end private applications. State continuity, however, has a wallet corollary: the user must retain, reconstruct or securely rescan the state that gives private chain objects meaning.
The mainnet’s genesis block was created on March 17, 2026, according to Midnight’s March network report. That report names Lace, CTRL and 1am as primary wallet interfaces for private smart contracts and assets. Their recovery design is therefore not a secondary user-experience detail. It defines whether private ownership survives ordinary events: a broken phone, an obsolete laptop backup, a wallet migration or a long interval offline.
The chain has evidence, not a readable balance
A shielded payment normally leaves several distinct artifacts.
A commitment is a public chain object that commits to a private note, such as an asset amount and ownership information, without revealing those values. A nullifier is a public value revealed when that note is spent. It prevents the same note from being spent twice without identifying the note directly. The wallet must also hold the private material needed to recognize notes addressed to it and, later, create a valid spend.
The chain can therefore establish two public facts: a commitment was created and a nullifier was not previously used. It cannot simply return “Alice owns 42 units” to a generic query. That query would defeat the privacy model.
A restored wallet needs to derive candidate keys from the seed, obtain the relevant public chain history, inspect encrypted outputs, determine which outputs it can decrypt, reconstruct its spendable note set and verify that none of those notes has been nullified. Private contract execution can add witnesses, encrypted state and synchronization metadata to the same process.
The important distinction is between data that is public but opaque and data that is private but reconstructable. Commitments and nullifiers can be fetched from the network. The seed can derive deterministic key material if Midnight’s wallet design exposes the appropriate derivation path. But a wallet still needs either an encrypted record of private state or the ability to scan history and recreate it.
A provider that calls a seed phrase a complete backup without explaining this boundary has described key recovery, not necessarily state recovery.
Four recovery cases, four different risks
The simplest case is a new device with a current encrypted wallet backup. The backup may contain recognized notes, encrypted contract state, synchronization checkpoints and local metadata. The device unlocks it with a password or recovery key, checks chain progress from its stored checkpoint and then scans forward. This is fast, but the backup becomes a concentrated target. If copied in plaintext, it may expose balances, counterparties, application interactions and a spend path.
The second case is seed-only restore. The wallet derives its keys and scans from a safe start height, potentially genesis. This minimizes dependence on provider-hosted backups, but moves the cost to private scanning. The device must retrieve enough public history to identify its encrypted outputs. A server that observes which blocks, filters or notes a client requests may infer timing, wallet age or application use even if it cannot decrypt the payloads.
The third case is an old backup. It includes notes known at backup time but misses later receipts, spends and private contract changes. A correct wallet cannot treat its local note list as final. It must replay public state after the checkpoint, identify new encrypted notes and test whether old notes were nullified. Otherwise it risks showing a stale balance or attempting a spend that the network rejects.
The fourth case is migration between wallet providers. Here recovery depends on portability. Can Wallet B derive the same addresses? Can it import encrypted private state from Wallet A? Can it perform a full scan using compatible viewing material? Can it reproduce the contract witnesses necessary to interact with a deployed application? A private system that supports only provider-specific backups may preserve cryptography while creating practical custody lock-in.
Local state is part of the security boundary
Midnight’s FAQ is unusually direct about the consequence: loss of local private state can make shielded assets inaccessible. That suggests a useful architecture rule. Treat the local database as a security object, not a disposable cache.
Some state should be recomputable:
| State item | Publicly obtainable? | Safe to recompute? |
|---|---|---|
| Block headers and public history | Yes | Yes |
| Commitments | Yes | Yes |
| Nullifiers | Yes | Yes |
| Address derivation from seed | No, but derivable | Usually |
| Recognized encrypted notes | Not directly | Only by scanning or restoring backup |
| Spend witnesses and private contract data | Not necessarily | Depends on protocol and retained data |
| Synchronization checkpoints | No | Yes, but potentially expensive |
The difficult row is the private note and contract layer. A wallet must not assume that every item needed to spend can be recreated from a public node response. If private contract state is encrypted to the user and only stored locally, recovery requires a recoverable encrypted copy, a deterministic derivation procedure or a chain-history scan that can recreate it.
The word “witness” deserves precision. A witness is the private evidence a prover uses to show that a statement is true. In a shielded spend, it may establish membership of a note commitment in an authenticated structure without revealing which commitment is being spent. The proof crosses the protocol boundary. The witness should not.
No private witness transmitted
This has direct recovery implications. A cloud backup can safely hold only what it encrypts under key material the cloud provider cannot access and without any remote decryption or proof-production role. If the provider can access spending keys or help produce valid spends, it is not merely storing a backup. It has become a practical co-custodian.
Viewing keys solve one problem and create another
The FAQ also documents viewing-key disclosure. A viewing key can allow a third party to inspect selected private information without receiving full spending authority. That is useful for tax reporting, audit, customer support, delegated operations and institutional controls.
It is not a free privacy feature. A viewing key has scope. Wallet architects must define whether it exposes incoming payments, outgoing payments, asset types, transaction timing, balances, contract interactions or only a bounded subset. They must also decide whether disclosure is revocable in practice. If a recipient copies decrypted records, later revocation cannot make that copy disappear.
A recovery system should separate at least three credentials:
- Spend authority, which can create valid spends.
- View authority, which can decrypt or recognize defined private activity.
- Backup authority, which can restore encrypted state without automatically granting either of the first two outside the user’s device.
Combining all three in a seed-only export is simple but increases blast radius. Separating them supports selective disclosure but increases key-management complexity and recovery failure modes.
DUST delegation needs an explicit restoration story
Midnight uses DUST in its transaction model, and the FAQ describes DUST alongside the system’s public commitments, nullifiers and address mechanics. If a user delegates DUST to an application while the chain continues advancing, the wallet must track more than a static balance.
The recovery path should identify the delegation relationship, current authorization conditions, consumed resources, pending application actions and any revocation or expiry state. If those objects are private, an old backup cannot be presumed authoritative. The restored wallet must scan forward and reconcile its local view against public commitment and nullifier evolution.
This is also where application wallets can leak more than base wallets. A generic wallet may scan for owned notes. An application-integrated wallet may query narrowly for a particular contract, resource pattern or recent interaction. Those queries can expose access patterns to a remote service. Privacy is preserved at the proof layer only if surrounding synchronization, indexing and telemetry do not recreate the same history elsewhere.
Assisted recovery is a custody decision
There are several defensible recovery models, but each trades a different risk.
Seed-only recovery reduces reliance on a provider, but may impose a slow scan and may fail where unrecoverable private state was never backed up.
Encrypted wallet backups improve restore speed and preserve local state, but the encryption design, password recovery process and backup portability become security-critical.
Viewing-key recovery helps monitoring and reconciliation, but should not be confused with spend recovery.
Assisted recovery can let a service retain encrypted fragments or help locate state, but the protocol must prevent the service from learning a user’s transaction graph merely because it helps restore it.
Social or institutional custody can reduce single-device failure by splitting recovery authority among guardians or policy signers. It also adds availability risk, collusion risk and governance questions that cryptography does not eliminate.
The only external reaction supplied for this story is cautious rather than hostile. In a forum post, James, Bitrove’s founder, raised practical questions about wallet support, hardware-wallet compatibility, token availability and ecosystem maturity while supporting Midnight’s privacy-first escrow model. No opposing view was provided in the available material.
For wallet teams, the production test is concrete. Restore from a seed alone. Restore from an encrypted backup that is 90 days old. Restore after notes have been received and spent. Restore after a DUST delegation changes. Restore into a different wallet implementation. Record every network request during each test.
If the wallet restores funds but reveals the user’s entire transaction history to an indexer, recovery worked and privacy failed.
Jared Zimmerman is not a person. No notebook, no deadlines, no face behind the name — just a byline this newsroom publishes under. Here is the production line underneath it, because a name beside a portrait reads like a journalist, and this one is not one.
The models. Writing: gpt-5.6-luna and gpt-5.6-terra. Out on the live web: gpt-5.6-terra and gpt-5.6-luna. Pictures: gpt-image-1 and flux. Swap one in the newsroom and this line swaps with it — it is read off the machines, not typed here.
How a story is made
- Research. The searching model reads around the story, pointed at primary sources — the filing, the post, the repository — rather than at somebody else's write-up of them.
- Writing. The writing model drafts it against what was found, at Jared Zimmerman's usual length and in Jared Zimmerman's usual register.
- The loop. A reviewer reads the draft and sends it back with notes. Then reads it again. A piece can go round several times before it leaves the building.
- Enrichment. A quotation has to appear word for word on the page it is taken from. A chart may only use figures that appear in the source it cites. Whatever fails is dropped, and the reason is kept.
- Fact check. A last pass hunts for claims the article makes and its sources do not.
- A human stop. Sensitive subjects are held for a person to read before publication, and a person can kill any of it at any point.
If that sounds less like a newsroom and more like a factory: quite. It is called Press Factory.
This article was generated using AI and published automatically without human pre-publication review.
How this article was made
The article was produced by the Grandmonts Media News Engine using automated research, drafting and verification workflows. No human editor reviewed the article before publication. Grandmonts Media remains responsible for the published content. Errors can be reported at office@grandmonts.cz.