HFIPay: Privacy-Preserving, Cross-Chain Cryptocurrency Payments to Human-Friendly Identifiers
Jian Sheng Wang · 2026
Read HFIPay: Privacy-Preserving, Cross-Chain Cryptocurrency Payments to Human-Friendly Identifiers by Jian Sheng Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-26970.txt.
Introduction
Wang presents HFIPay, a protocol for paying cryptocurrency to identifiers like email addresses. Direct maps would expose balances and payment histories to anyone. A relay resolves the name off-chain and posts only blinded, per-payment data on-chain. The sender first verifies a quote that blocks recipient swaps by the relay. Claiming then needs a zero-knowledge proof of control over the hidden identity. The author defines games for enumeration and pre-claim unlinkability. Relay compromise and post-claim linkability are mapped too.
Catalog record
| CREATOR | Jian Sheng Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 69932 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.26970
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
HFIPay: Privacy-Preserving, Cross-Chain Cryptocurrency Payments to Human-Friendly Identifiers Jian Sheng Wang Abstract Sending cryptocurrency to an email address or phone number should be as simple as a bank transfer, yet naïve schemes that map identifiers directly to blockchain addresses expose the recipient’s balances and transaction history to anyone who knows the identifier. HFIPay separates private routing , sender-side quote verification , and on-chain claim authorization . A relay resolves the human-friendly identifier off-chain and commits only a per-intent blinded binding ρ i \rho_{i} plus the quoted payment tuple; the chain sees neither the identifier nor a reusable recipient tag. In a verified-quote deployment, the relay returns a sender-verifiable off-chain proof linking ρ i \rho_{i} to an attested binding-key commitment, so the relay cannot substitute a different recipient before funding. To claim, the recipient proves in zero knowledge—via ZK-ACE [ 2 ] —that the funded intent’s blinded binding matches a handle derived from the same deterministic identity, authorizing release of the quoted asset and amount to a chosen destination. We formalize two privacy goals: enumeration resistance and pre-claim unlinkability , and distinguish a baseline deployment (relay trusted for binding correctness) from the verified-quote deployment (binding is sender-verifiable without a public registry). When composed with an n-Vm runtime, the same mechanism extends to cross-chain settlement. The result is a relay-assisted but non-custodial architecture: relays are privacy and availability dependencies, but cannot redirect funds. Keywords: identifier-based payment, privacy-preserving, verifiable quote, blinded claim binding, zero-knowledge authorization, cross-chain 1 Introduction 1.1 Motivation Cryptocurrency adoption is hindered by a fundamental UX barrier: users must manage opaque hexadecimal addresses, select the correct network, and possess native tokens for transaction fees. These requirements make peer-to-peer crypto payments inaccessible to non-technical users and error-prone even for experienced ones. By contrast, human-friendly identifiers such as email addresses, phone numbers, and social media handles (e.g., X/Twitter usernames) are ubiquitous and already serve as primary anchors for online identity. If cryptocurrency could be sent to such an identifier—with no wallet setup, no address copying, no gas token acquisition—the resulting experience would be indistinguishable from traditional digital payments. 1.2 Privacy Challenge A naïve implementation maps user identifiers directly to blockchain addresses via a deterministic function (e.g., \mathsf{keccak256}(\texttt{identifier})\to\texttt{address} ), where the identifier might be an email, phone number, or social handle. This is simple but introduces a critical privacy vulnerability: any third party who knows a user’s identifier can derive their on-chain address and inspect their entire transaction history, token balances, and counterparty relationships. For public figures, this is especially dangerous. An adversary knowing only [email protected] , a public phone number, or a well-known X handle could reconstruct Alice’s complete financial profile on every supported chain. 1.3 Our Approach HFIPay solves this by decoupling identifier routing , recipient binding , sender-side quote verification , and claim authorization . The sender specifies only a human-friendly identifier such as an email address. The relay privately resolves that identifier to a registered recipient record, samples a random intentId , derives a one-time deposit address, and commits only an intent-specific blinded binding \rho_{i} together with the quoted payment metadata on-chain. In the verified-quote deployment, the relay must additionally provide the sender with an attested quote proving that the quoted \rho_{i} is tied to the intended recipient’s hidden binding handle. The recipient later claims by proving control of the deterministic identity bound to that blinded value. The key properties are: 1. Enumeration resistance. No on-chain data reveals which funded intents belong to a known email, phone number, or social handle. 2. Pre-claim unlinkability. Multiple pre-claim payments to the same identifier do not expose a reusable public recipient tag. 3. Post-quote claim correctness. Once a sender has accepted a valid quote and the resulting intent stores the quoted blinded-binding tuple, claim authorization is enforced on-chain by a zero-knowledge proof rather than by relay discretion. 4. Device-portable recovery. The recipient can rederive the same deterministic identity across devices using Va-Dar , avoiding a public identifier registry. 5. Crypto-invisibility. In an integrated client flow, neither sender nor recipient needs to handle chain-specific addresses or gas manually. 6. Cross-chain routing. Combined with an n-Vm architecture, the same claim mechanism can release funds onto a different destination VM. Informally, HFIPay aims to provide the UX of “send to email” while keeping identifier resolution private and moving the spending authority check onto the chain. 1.4 Contributions This paper makes the following contributions: 1. Protocol design. We define HFIPay as a relay-assisted routing layer with intent-specific blinded bindings and quoted asset tuples committed before funding and an optional sender-verifiable quote, so that recipient substitution can be detected before funds are sent and post-funding claim authorization is enforced on-chain (Section 3 ). 2. Authorization composition. We instantiate claim-time authorization with ZK-ACE [ 2 ] and show how Va-Dar [ 3 ] supplies private recipient recovery/discovery across devices without public identifier enumeration (Sections 2 and 3 ). 3. Privacy analysis. We formalize enumeration resistance and pre-claim unlinkability against on-chain observers, and explicitly characterize what becomes linkable after claim (Section 5 ). 4. Binding modes. We distinguish a baseline deployment from a verified-quote deployment that uses attested off-chain binding proofs rather than a public identifier registry (Sections 2 and 6 ). 5. Cross-chain and implementation sketches. We show how the same claim relation composes with an n-Vm runtime and outline compatible EVM and Solana realizations (Sections 4 and 7 ). HFIPay builds on the Ace-Gf identity framework [ 1 ] , uses ZK-ACE for chain-verifiable authorization [ 2 ] , and uses Va-Dar for private deterministic recovery [ 3 ] . Its routing mechanism is related to stealth-address systems (ERC-5564 [ 4 ] ) but operates at the application layer with a different trust boundary; a detailed comparison appears in Section 8 . 2 System Model 2.1 Participants The protocol involves five roles: Definition 1 (Sender) . A party A who initiates a payment by specifying a recipient email address e_{B} , an asset identifier a , an amount v , and a target chain identifier c . Here a denotes the canonical source-chain asset descriptor, e.g., native ETH/SOL, an ERC-20 contract address, or an SPL mint. If refund is enabled, A also authorizes a refund destination \gamma_{A} using a sender-controlled signing credential or smart-account policy on the relevant chain. Definition 2 (Recipient) . A party B who controls a deterministic identity root \mathsf{REV}_{B} . From this root, B can derive an identity commitment \mathsf{ID_{com}}_{B} as in ZK-ACE [ 2 ] , and for each binding epoch e a private claim-binding handle u_{B,e}=H(\mathsf{Derive}(\mathsf{REV}_{B},\mathsf{Ctx}_{\mathsf{bind}}\|e)). B also derives the corresponding off-chain binding-key commitment K_{B,e}=H(\texttt{"hfipay:bind-key"}\|u_{B,e}). Deployments SHOULD choose e from a coarse public epoch schedule (e.g., a daily or weekly rotation counter) rather than a recipient-specific random label, so that revealing e as intent metadata does not itself create a reusable recipient tag. B may load or recover the same root across devices using Va-Dar [ 3 ] . Definition 3 (Relay Service) . A service \mathcal{R} that maintains a private directory \mathsf{Norm}(e)\mapsto(\mathsf{ID_{com}},u_{e},K_{e},e,\texttt{mode},\texttt{metadata}), and a private mapping from identifiers to pending intents. \mathcal{R} pays blockchain transaction fees on behalf of users (relayer-pays model). Definition 4 (Binding Layer) . An optional application-layer or third-party service \mathcal{I} that helps establish the binding between a