ZENIP Idea: Quantum-Resistant Finance with Privacy on Base

Abstract

The proposal in one line: Horizen becomes the leading application for quantum-resistant finance with privacy on Base through two complementary pieces deployed directly on Base: a shielded pool for private transactions and balances, and Vela as the confidential-compute layer that makes private decentralized finance programmable.

Similar privacy systems are already in operation in the Ethereum arena, and this Idea proposes for the system to be built with regulatory compliance in from the start: screened deposits, selective disclosure, and non-custodial contracts.

This ZENIP proposes a direction, not parameters. The proof system, fees, staking mechanics, and timelines are deliberately left to a technical feasibility study and a governance vote, and the open items are listed below rather than guessed at.

The intention is to ship a Minimum Viable Product as soon as possible, then continue to grow and expand the capabilities over time, shipping additional products and features as they are developed.

The two parts of the direction

1. Shielded pool on Base — private transactions and balances.
Contracts on Base that let users shield assets, hold them privately, transfer them, and unshield. Notes are designed for a quantum future. Deposits are screened. Users can prove specific payments to auditors or counterparties without making everything public. This is the privacy layer for value.

2. Vela on Base — programmable private decentralized finance.
Vela is the confidential-compute environment that can operate over private data the pool alone cannot. A shielded pool proves facts about one user’s notes; it cannot compute across multiple users’ private state. Vela can, which is what enables private applications: private order books or matching, confidential settlement logic, private lending terms, agent-to-agent negotiation, and other DeFi functions that need to stay encrypted until the final attested result. Vela is deployed and improved on Base alongside the pool, not as an afterthought. It opens as a marketplace of attested operators; bonding and slashing rules come from the technical study. It is a service market, not a requirement for basic pool transfers to function.

Together they are the system: the pool for private value, Vela for private computation on that value.

Candidate early use cases

Several uses are realistic. We list them so the community can push on priority:

Pool-first uses

  • Private B2B settlement between known counterparties (supplier payments, partner settlements, inter-company transfers). Parties usually already know each other; the pool mainly hides amounts and relationships from the public chain. Selective disclosure fits naturally.
  • AI-agent private transfer. Agents paying for data, compute, and inference. Docs and interfaces optimized for Agentic interaction.
  • Private holding and yield. Shield assets, earn yield, unshield later. Privacy is mainly about balances.
  • Private swaps inside the pool. Enter with one asset, trade, exit without revealing path or size.
  • DAO / protocol treasury management. Operating funds kept shielded.
  • Private donations or grants.

Uses that need Vela (or benefit strongly from it)

  • Private DeFi applications. Confidential matching, private lending parameters, sealed-bid style flows, or other logic that must compute over more than one user’s private data.
  • AI-agent private operations. Agentic onchain DeFi without broadcasting strategy; Vela can hold the private computation. The goal would be for an AI agent to be able to access and use a Vela enclave on its own.
  • Selective, attested compliance or settlement logic that should not be fully public.

Proposed launch focus for the pool: private B2B settlement between known counterparties. It can bring repeated commercial volume and has a clear privacy rationale. Vela development and deployment on Base proceed in parallel so that programmable private finance is ready as the pool gains usage. Other uses remain available and can be emphasized as the system matures. The technical study and community input can reorder priorities.

Applications on Base for Private Payments and Private Finance

Ship in atomic steps. Working products ship one at a time to generate activity and feedback, instead of waiting for a complete system. Pool and Vela work advance together:

  1. Private transfers and holding (pool MVP). Shield stablecoins on Base, hold them privately, and send to any address. Early usage aimed at B2B settlement between known counterparties. Deposits screened against sanctions lists. Open payroll or broad consumer payments are not the launch focus.
  2. Privacy you can prove. Selective disclosure: prove a specific payment to an auditor, tax authority, or counterparty without publishing it, plus the documented screening process. Privacy and demonstrable compliance arrive together.
  3. Vela on Base, usable for private computation. Vela deployed (or expanded) on Base so builders can run confidential jobs. Initial operation can be more centralized; the path to a marketplace of attested, ZEN-bonded operators is explicit. This is what turns the pool into programmable private finance.
  4. Staking for the system. ZEN has an existing staking program. This new direction enables the creation of an updated staking program with mechanics chosen to support anonymity-set growth, long-term participation, and (where appropriate) Vela operator bonding, while minimizing structures that invite securities scrutiny. Exact design set by the study.
  5. Private movement across assets and deeper private DeFi. Move shielded value between supported assets without publishing the path. Yield and other applications that combine the pool and Vela follow.

Each step is a shipping product with users, not a research milestone. Which study results could reorder the steps is exactly what the feasibility study and the open items below settle.

Specification (high level, parameter neutral)

  1. A shielded pool on Base. Contracts deployed directly on Base, where stablecoins, wallets, and on/off-ramps already are. No separate chain and no bridge. The contracts are non-custodial and permissionlessly usable. Operating functions (deposit screening, default relaying, proof aggregation) sit with the Foundation.
  2. Post-quantum from the start. Notes encrypted with post-quantum cryptography, with post-quantum proofs in the initial pool deployment. The exact proof system will be selected by the technical feasibility study.
  3. Compliance people can use. Deposits screened against third party standard sanctions lists as a named, documented, and logged process. Users can generate proofs of specific payments for auditors or tax authorities. Optional exit-side tools (such as proofs of clean origin) will be evaluated. Procedures for post-deposit issues will be published.
  4. Foundation-run proof aggregation at launch, so verification on Base stays cheap and reliable. The output is a proof anyone can re-verify on Base.
  5. Vela as the programmable private-finance layer on Base. Deployed and improved on Base in parallel with the pool. It enables confidential computation over private data for DeFi applications. It opens as a marketplace: attested operators run hardware and bond ZEN to qualify for jobs and back provable failures. Exact bonding and slashing rules come from the technical study. Basic pool transfers do not depend on Vela; private applications that need multi-party private compute do.

What ZEN does. ZEN has two jobs, both named in the proposal.

First, ZEN staking is an intention for the design. Options will be evaluated to support pool security and liquidity (including cold-start incentives that grow the anonymity set) and Vela participation, while avoiding fee-sharing structures that create clear securities risk.

Second, ZEN-bonded compute for Vela: operators bond ZEN to qualify for jobs and back provable failures, with the rules set by the technical study. One design principle up front: nobody needs to hold ZEN to make a payment in the pool.

Still being worked out, by design:

Proposed direction Still open, set by study or vote
Shielded pool on Base for private transactions The proof system and exact cryptography
Vela on Base for programmable private DeFi Deployment timing, marketplace rules, and ZEN bonding roles
Private B2B settlement as the proposed pool launch focus Fees, pricing, onboarding, and whether another early use case should lead
ZEN staking Staking mechanics, cold-start incentives, and reward design
Post-quantum design from the start Which parts launch first and which follow

Motivation

Businesses that settle with suppliers or partners in stablecoins today publish every amount and relationship on a public ledger. That leaks pricing, volumes, and commercial strategy. At the same time, decentralized finance itself is limited if every position, bid, or term must be public.

Banks already give customers private payments under rules that block prohibited transactions. On-chain finance needs both private value transfer and the ability to compute privately. For people and businesses to use onchain finance, it needs to provide better services than banks.

The shielded pool addresses the first. Vela addresses the second. Together they are how Horizen proposes to deliver quantum-resistant private finance on Base.

Horizen began in 2017 as ZenCash, a privacy coin, and Horizen moved away from privacy over the years because the granularity to create regulatory compliant private transactions and decentralized finance did not yet exist.
This is a proposal to use the 10 years of technological improvements to implement the original vision: private balances and payments, encrypted to resist quantum attack, provable to the parties who need to see them, and programmable through confidential compute.

ZEN continues to be a token on Base with no inherent privacy features of its own. What this proposal develops and deploys are applications and smart contracts on Base, within the Ethereum ecosystem, that give stablecoins and the other assets in the system practical privacy.

Background / Context

The Horizen Layer 3 is being wound down on Base’s schedule. Base’s Denim upgrade (November 5) makes Base far faster than a Layer 3 built on it can keep up with; when the upgrade reached Base’s testnet on October 21, the Horizen testnet stopped within the hour, and the chain’s operator has confirmed it cannot be patched in time.

ZEN on Base and on exchanges is not affected, and a separate announcement covers the wind-down, the exit window, and how remaining funds are returned.

What Horizen brings to the next step: ZEN and a community that has stayed through everything, Vela (a confidential-compute environment already in builders’ hands), and a cryptography team at Horizen Labs that has spent years on zero-knowledge and post-quantum work. The proposal puts those strengths directly on Base, as contracts that keep working through Base’s upgrades, where the users and the dollars already are.

Process and Reporting

The order is set: Special Council review, then open community discussion (this thread), then the Horizen Labs technical feasibility study, then a detailed dated roadmap published from the study’s results, then a ZEN-holder governance vote. Phases move forward when their conditions are met, not on a calendar.

Progress updates will be posted in plain language in this thread as the study reports. The first clarity on cryptography and costs is expected from the study itself, and the dated roadmap only after it.

Security & Privacy

Audit plans and the selective-disclosure tooling (proving a specific payment to a specific party) will be published as the implementation work locks each piece down. The design principle is that privacy and provability ship together: no product releases that force users to choose between confidentiality and compliance.

Operating functions are run by the Foundation with published methodology and logs sufficient to demonstrate that screening occurred. The underlying contracts remain non-custodial and permissionlessly usable.

Known Risks

Naming these ourselves, because each one is the reason behind a design choice, not a footnote.

  1. Post-quantum cryptography may not hit its numbers. Post-quantum proofs are bigger and slower than today’s, and the feasibility study may find they do not fit a fast payment product.
  2. Node operators for the pool could recreate the problem of third party dependency. The initial design of the pool’s plumbing (relaying, aggregation, coordination) is Foundation-run with published, verifiable output and permissionless fallbacks. ZEN bonded shielded pool nodes may be added as a future update. Vela operators are a marketplace, not pool-critical infrastructure.
  3. ZEN staking design. Staking should help grow the anonymity set, support Vela participation, and give holders a long-term role, while avoiding structures (especially ongoing fee-sharing presented as yield) that create clear securities risk. Mechanics will be set after the study.
  4. A new pool starts with a small anonymity set. Privacy is only as strong as the number of users in it, and day one numbers will be small. Focusing first on repeated B2B settlement is intended to bring real volume. Vela applications can add further usage as they ship.
  5. ZEN depends on Base. The L3 taught us what depending on someone else’s chain costs, and no L2 choice removes dependency entirely. What changes this time: the design is contracts, not a chain. Contracts ride Base’s upgrades instead of breaking on them, the note tree lives on Base where anyone can read and verify it, and withdrawals require nobody’s permission. The dependency becomes “we upgrade when Base does” instead of “we can be switched off.”
  6. Vela relies on outside hardware. It runs on AWS Nitro Enclaves today, so its confidentiality and availability depend on that provider
  7. Legal and regulatory review. Payments and token design raise legal questions that vary by jurisdiction. Legal analysis will be obtained as necessary, and staking designs will favor ZEN being locked and returned over ongoing claims on protocol fees.
  8. Sanctions and AML effectiveness of the screening model. Deposit screening is a necessary start. Limitations are real. Additional tools such as optional exit-side proofs of origin and published post-deposit procedures will be evaluated.

None of these are reasons to wait. They are the reasons the order is: study first, smallest useful product first, pool and Vela advancing together, verifiable plumbing before trusted roles, clear separation of protocol from operating functions, and updates in public.

What we are looking for from this discussion

  • Push on the direction. What breaks if we build this and it’s wrong? What did we miss?
  • On the two parts: Is the balance between the shielded pool (private transactions) and Vela (programmable private DeFi) right? What should ship first in each?
  • On the pool launch focus: Is private B2B settlement between known counterparties the right first emphasis, or should another use case lead?
  • Business owners: what would you actually need to run private supplier, partner, or inter-company payments through a shielded pool? What selective-disclosure features matter?
  • Builders: what private DeFi applications would you run on Vela attached to a shielded pool?
  • Community: what do you think, what are we missing here?

Rolf Versluis and Rob Viglione, cofounders of Horizen @finpunk tagged for visbility

7 Likes

Could we make third-party integrations a core part of this proposal? Instead of relying primarily on users coming to Horizen’s own apps, we could offer the shielded pool and Vela as privacy infrastructure that existing wallets, payment platforms, and DeFi protocols can integrate directly.

2 Likes

Fully behind this, I already wrote a whole post about why, so I’ll just throw out two things I didn’t see covered. First, why not have the DAO treasury be the first big user of the pool? Shield part of the treasury on day one and run grants and contributor payments through it. It helps the cold start since there’s real volume from the start, and it’s the best proof we actually trust our own product. Second, Vela running only on AWS Nitro is a single point of failure, and we literally just got burned by depending on one outside party. Is supporting other TEEs (Intel TDX, AMD SEV) part of the study? Even just on the roadmap would be nice. Also, the AI agent payments use case is kinda slept on imo. Agents paying for compute and data privately could end up being more volume than B2B, worth keeping an eye on.

5 Likes

yes, it will be open. Would love to have builders use shielded pool and Vela to build their own apps.

1 Like

these are all great ideas!

Rolf and I co-wrote this, so I fully support it. Three reasons:

It’s what Horizen was built for. We started in 2017 as a privacy project. The tools to do privacy that people and businesses can actually use, with screening, selective disclosure and cryptography designed for a quantum future, didn’t exist then. They do now and we’re perfectly positioned to take advantage of it.

It’s built where the users are. The L3 taught us what it costs to depend on another chain’s schedule. Contracts on Base ride its upgrades instead of breaking on them, and they sit next to the stablecoins, wallets and on-ramps people already use.

Study first, then ship in steps. Horizen Labs is committed to doing the feasibility work properly and publishing what we find, including where the numbers don’t work. The roadmap comes from that, and we can move fast.

There are some strong ideas already in this thread. Opening the pool and Vela to third-party wallets and apps, running DAO treasury payments through the pool from day one, and supporting TEEs beyond Nitro are all worth folding into the study. Keep them coming.

6 Likes

As the technical feasibility study takes shape, what is the envisioned trust assumption for Vela’s initial operator marketplace—will it rely primarily on cryptographic staking (ZEN bonds) and remote attestation validation, or will there be developer SDKs available for early builders to spin up custom execution enclaves during the Beta phase?

2 Likes

Love it. I just have one question: Circle can freeze the whole pool. All shielded USDC sits in one contract. Circle froze USDC at Tornado-linked addresses in 2022. One blacklist entry freezes every user’s funds. What happens to all shielded USDC if Circle blacklists the pool contract?

3 Likes

@Trololino I think this question is better rewritten as a requirement that the pool has to be able to survive this kind of shock.

2 Likes

Awesome question. We don’t know yet until we’re done with a series of research spikes, but I’d love to see ZEN bonds for Vela operators.

I think it needs to be able to automatically freeze funds marked as blacklisted funds. Good example is this Binance hack few days ago.

1 Like

There are some guidelines for things to do for Horizen Foundation to not be classified as a money transmitter. Here is a short list of things to consider. We already have a more extensive legal guideline document from Horizen Foundation legal counsel:

  1. CASP perimeter: Running the relayers and aggregating proofs does not make the Foundation a crypto-asset service provider, as long as it never holds keys or funds (not even in transit), the plumbing stays open-source and replaceable by anyone, it applies published rules mechanically, it charges no service fee, and the Foundation does not bolt listed services (swaps, ramps, managed accounts, etc.) onto the pool under its own name. The opinion sets these out as six design fences. One thing to watch: the ZENIP says “operating functions sit with the Foundation”, which is exactly the phrase a supervisor would pick on. The same post also says the contracts are non-custodial with permissionless fallbacks, and that is the description to use everywhere instead.

  2. Anonymity-enhancing coin: on the text of the AML Regulation, the better view is that ZEN is not one, because the privacy feature is not built into the token. In practice, assume that banks and exchanges will treat it as if it were. It does not matter though, provided that the pool lets obliged entities identify the holder and the transaction history. MiCAR already says so for trading venues in Article 76(3), and the AML Regulation’s prohibition targets the institution’s own accounts, not a customer’s history. The opinion lists the six things a CASP must be able to do and how the pool should make each of them easy.

  3. Reporting capability: it is not a nice-to-have, it is the condition of access. The travel rule, DAC8 (live since January, first reports due next year) and Article 76(3) each need it. One option is a two-layer model: a scoped, revocable disclosure key, plus a standing authorization the user gives a CASP at onboarding so the CASP can pull reports for its ongoing monitoring without asking each time. That is the shape supervisors already accept in neighboring regimes.

  4. Screening: the Foundation should own and publish the policy, an independent named analytics vendor should perform the screening, the contract should enforce it, and a third party should review it annually. Screen against the EU, UN, OFAC and UK lists, re-screen the association set daily, and keep tamper-evident logs for five years. The opinion has the log schema that I envision would work out the best and easiest for this kind of use-case.

4 Likes

I’ve been on the fence with stuff like this over the years, especially at the L1 level, but for what we’re doing with Horizen I agree @Trololino. We do the community and our industry no favors by enabling an environment where criminals (and state actors) have an easy time stealing boat loads of money. We should make it as hard as possible to block them and as easy as possible for the community to recover funds when the inevitable incident happens.

3 Likes

Strong support for this direction. Privacy is where Horizen started, and building it directly on Base, with no bridge, removes most of the friction that hurt earlier attempts.

A few points from someone who has spent a lot of time on post-quantum privacy on Base:

1. Make the permanent parts post-quantum from day one. The note format, commitments and nullifiers can’t be changed after launch without migrating every user’s funds. The proof system can evolve later. These can’t. That is where “quantum-resistant” has to be true on day one.

2. On-chain verification on Base is feasible. From my own testing on Base Sepolia, a fully post-quantum private transfer can be verified natively in a single Base transaction, with no trusted setup and no batching. Proving on the user’s device takes seconds. The feasibility study can aim high here: it doesn’t have to start with a quantum-vulnerable wrapper and replace it later.

3. Compliance through receipts the user holds, not a master key. Users should be able to prove a specific payment, or where their funds came from, to an exchange, an auditor or a counterparty, without exposing anything else and without a global key that sees everything. That is what makes the pool usable for exchanges and on-ramps, especially in the EU.

4. On @Trololino’s point about a single-issuer freeze. It is a real risk and it should be a design requirement, but the pool can’t fully remove it: if an issuer blacklists the pool contract, that asset can’t leave the pool, whatever the pool’s own rules say. What the design can do:

  • Give issuers and regulators no reason to act against the pool. Screening at entry and receipts users can share mean the pool can show it is not a channel for sanctioned funds.
  • Bound the damage. Use per-asset caps, and isolate assets so that a freeze of one doesn’t touch the others.
  • Don’t add freeze powers of its own. No key should be able to pause exits or freeze individual notes. A pool that can freeze users’ funds is a bigger regulatory and trust problem than the one it solves.
  • Be honest with users. Holding USDC in the pool carries the same issuer risk as holding it anywhere else.

5. Start narrow, expand later. Private stablecoin transfers first. Then cross-asset features such as private swaps and yield, once each has had its own review. These are technically achievable, but each one adds regulatory and security surface.

6. Integrations from day one. +1 to @Jimmy. A pool only wallets can’t talk to will stay empty. A clean client SDK and an open relayer interface should be part of the first deliverable. Horizen Wallet would be glad to be one of the first integrations.

Looking forward to the feasibility study.

2 Likes

I really like the direction of this proposal, especially the shift from building our own blockchain infrastructure to delivering privacy applications directly on Base.

One suggestion I’d like to make is to prioritize distribution and third-party integrations from day one, alongside technical development.

We’ve spent years building infrastructure and trying to attract users and liquidity into our own ecosystem. This time, why not bring Horizen’s privacy technology directly to where the users and liquidity already are?

For example, we could make the Shielded Pool and Vela easy for existing wallets, payment platforms, and DeFi protocols like Uniswap or Aerodrome to integrate through SDKs, APIs, and adapters.

I’d also suggest identifying a few real B2B stablecoin payment customers early in development, so we’re building around actual demand rather than assumptions.

Finally, it would be great to establish clear adoption metrics, such as active integrations, paying customers, transaction volume, and protocol revenue.

The goal shouldn’t just be to build great privacy technology. It should be to make Horizen’s privacy technology something existing businesses and applications actually want to use and pay for.

I believe this could make a real difference in achieving sustainable product-market fit.

4 Likes

Agreed with @finpunk here and thanks @Trololino for raising it. Surviving an issuer freeze should be a design requirement.

Base’s sequencer is a similar risk, so one suggestion for the technical study:

Base’s terms let Coinbase prohibit specific wallet addresses from transacting through its sequencer. The fallback is forced inclusion through L1, which today can take up to 12 hours. Base’s docs don’t describe it and I could not find anything saying whether it stays the same after the remaining hardforks as Base moves off the OP Stack. It is worth adding to the study: confirm and test that a user can withdraw from the pool through L1 even if the sequencer refuses their address or anything touching the pool, including whether a post-quantum withdrawal proof fits within the gas cap on L1 deposit transactions and stays affordable on L1. If that passes, “withdrawals require nobody’s permission” will be true against Base’s possible censorship risk as well.

Some other thoughts:

Exchanges

While the ZENIP is clear that ZEN itself has no privacy features, exchange listing teams will still connect “Horizen” with “privacy pool.” I would recommend briefing exchanges that list ZEN before the pool goes live, so they hear it from Horizen first, and working through any concerns with them, particularly what they need before accepting a withdrawal from the pool.

DAO treasury as first user

+1 to @rocket’s idea. Treasury spending has to stay accountable to ZEN holders, which makes it a good fit for selective disclosure: shield grant and contributor payments and publish a regular disclosure report to holders using the pool’s own tools. That makes the treasury a live demo of provable privacy.

Minimum standards for integrations

+1 to @Jimmy on integrations from day one. I would suggest a published list of requirements before any wallet or front end gets listed or promoted though, things like: open source and audited; self-custodial, with keys that never leave the user’s device; clear about what the operator can see.

3 Likes

Solid feedback @Vano, thank you. I really like the withdrawals without anyones permission concept, given where Base sits. Very cool idea to chase down.

1 Like

It would be interesting to know if the Privacy pool can transact with Tokenized assets, bonds and stocks?

1 Like

For sure at some point and we’ll have several options,

  1. An adapter to other DeFi apps, like what Railgun does
  2. Private finance in Vela, linked to the shielded pool
  3. Potential for apps to build directly with the shielded pool (fuzziest at the moment)
1 Like

Strong support for this direction. Splitting the work between a pool for private value and Vela for private computation is the right call, because each covers what the other can’t.

A shielded pool on its own can hide balances and transfers, but it can’t run logic across many users’ private positions. Vela can, but on its own its entry and exit points are public. Today, depositing into a Vela app shows the wallet, token and amount on Base, every request carries the app ID in plain view, and withdrawals show the recipient and amount. Linking the two closes that gap: value stays shielded on the way in and out, and Vela does the private computation in between.

Private lending is a natural early proof of this. Lending needs exactly what the pool alone can’t provide: shared liquidity, interest across all positions, health checks against live prices, and liquidations. All of that has to run over many users’ private data at once. We’re building a lending app on Vela (NoctFinance), and the pool would let users enter and exit it without leaving a public trail, which is the missing piece today.

A few things that would make the link work well for any Vela app:

• Shielded entry: move value from notes into a Vela app with the amount carried only inside the encrypted request, not on Base.

• Shielded exit: Vela pays withdrawals, refunds and settlements back into notes rather than to a public address.

• One shared anonymity set: funds enter Vela in general and move between apps inside the enclave, so each app isn’t individually visible on-chain. This also helps with the cold-start problem, since every Vela app adds to the same crowd.

• Compliance that connects: pool screening and user-held payment receipts that apps can link to their own reporting, with no global master key.

• An early testnet: a Base Sepolia pool and SDK during the feasibility study, so builders can test the integration and feed results back.

+1 to opening this to third-party builders from day one. We’d be glad to help test the pool-to-Vela path once there’s something to try.

Daniel from Noctfinance

3 Likes