Aztec and Railway — Ethereum Privacy Layers
How Aztec's zkRollup and the RAILGUN shielded pool (used through the Railway wallet) approach privacy on Ethereum from different architectural angles, and what shipping ZK-based Ethereum privacy actually looks like.
The basics
Privacy on Ethereum's Public Chain
Ethereum is fully transparent by default — every transaction, balance, and interaction is public. This transparency is core to Ethereum's audit-and-verify security model but hostile to privacy needs (business confidentiality, personal safety, tax planning). Multiple projects are building privacy layers on top of Ethereum. Aztec provides shielded transactions in a dedicated privacy rollup. RAILGUN, used through wallets such as Railway Wallet, provides privacy for existing Ethereum and EVM tokens through a zk-SNARK shielded pool. Each represents a different architectural bet on Ethereum privacy.
Aztec's Rollup Approach
Aztec is a ZK rollup dedicated to private smart contracts and transactions. Users deposit assets from L1 into Aztec; within Aztec, they can conduct private transactions using ZK proofs. Withdrawal back to L1 reveals the withdrawn amount but not the intermediate history. Aztec's ZK proof system (based on PLONK derivatives) is optimised for privacy applications. The trade-off vs generic rollups: less general-purpose, more privacy-optimised. Aztec has pivoted several times through its history; current architecture is the third major iteration.
RAILGUN's Shielded-Pool Approach
RAILGUN, accessed through wallets such as Railway Wallet, takes a different approach — smart contracts on Ethereum (and other EVM chains) that enable shielded transfers of existing tokens. Users deposit into RAILGUN's shielded pool; transfers within the pool are private (like Tornado Cash mechanics but for arbitrary tokens); withdrawals reveal recipient and amount. RAILGUN's advantage: works with existing tokens without requiring bridge. Disadvantage: privacy is only within the pool, exit reveals information. Both models have legitimate uses; users choose based on specific privacy needs.
- Aztec: dedicated ZK rollup with private smart contracts
- RAILGUN (used via wallets such as Railway): privacy layer for existing tokens via shielded pool
- Different architectural bets on how to add privacy to Ethereum
- Multiple production deployments emerging post-Tornado Cash
Key Takeaways
- Ethereum's transparent design is hostile to privacy; privacy layers are needed
- Aztec's rollup approach: private transactions within a dedicated environment
- RAILGUN's approach (via wallets such as Railway): shielded pool for existing tokens
- Both are alternatives to mixers like Tornado Cash (delisted by OFAC in March 2025) with different trade-offs
Going deeper
Aztec Contract Programming
Aztec smart contracts use Noir, a Rust-inspired language designed specifically for zero-knowledge computation. Contracts can have both public state (visible on Aztec's L1 settlement) and private state (visible only to authorised participants via ZK proofs). This mixed model enables applications like private DeFi (transactions private, protocol state public), private auctions, and confidential financial products. Building on Aztec requires understanding both the ZK circuits and traditional smart contract logic. The developer experience is more complex than pure Solidity but tractable.
Aztec's Multiple Iterations
Aztec's journey has been iterative. Aztec 1.0: initial rollup, launched 2021, wound down. Aztec Connect: bridge-based privacy layer, launched 2022, sunset in 2024. Aztec Network: current rollup with Noir; its Ignition Chain went live in November 2025, transactions were switched on in the Alpha network on 31 March 2026, and the Aztec Tierra (Alpha V6) upgrade followed on 3 November 2026, with a production-ready Beta V1 on the roadmap for around Q2 2027. Each iteration attempted different architectural approaches. The current Aztec Network aims to be a general-purpose privacy rollup with meaningful DeFi ecosystem. Whether the third iteration achieves adoption remains to be seen; the technical foundations are stronger than earlier iterations.
Railway's Compliance Focus
Post-Tornado Cash sanctions, Railway and RAILGUN have emphasised compliance. They offer 'Private Proofs of Innocence' (Privacy Pools) features — users can generate ZK proofs demonstrating their funds don't derive from sanctioned addresses. This provides selective disclosure for regulated interactions (exchanges, KYC'd businesses) while maintaining privacy for other purposes. The design is a response to the Tornado Cash regulatory environment: privacy protocols that are structurally difficult to sanction. Whether the approach is legally robust is still being tested.
The Post-Tornado Privacy Landscape
Multiple privacy projects have emerged post-Tornado sanctions, each with different compliance postures. Some emphasise pure privacy (regulatorily risky); others emphasise compliance (adoption risk from privacy-focused users). The middle ground — compliant privacy with proof-of-innocence — is where several major projects are converging. Whether this middle ground satisfies regulators or users adequately is still being tested.
Key Takeaways
- Aztec contracts use Noir language; mixed public/private state model
- Aztec has had multiple iterations; current Aztec Network is most mature
- Railway emphasises compliance via proof-of-innocence features
- Post-Tornado landscape has convergence toward 'compliant privacy' middle ground
Under the hood
ZK-based Privacy Cost Analysis
Privacy has real costs. Proving costs: ZK proofs are computationally expensive to generate; user devices sometimes struggle. Gas costs: shielded transactions consume substantially more gas than transparent equivalents. UX costs: privacy adds complexity (wallet management, note management, proof generation waits). Composability costs: private assets are less composable with public DeFi. These costs limit adoption. Privacy is genuinely valuable but comes at practical costs users must weigh against their specific needs.
Client-Side Proving
Modern ZK privacy systems increasingly generate proofs client-side (in users' browsers or apps). This preserves privacy — no server sees the private state. But client-side proving means users' devices must handle the compute; complex proofs can take minutes on typical hardware. WebAssembly-based provers (used by Aztec, others) have improved this substantially. Mobile proving is now feasible for simple transactions. Complex proofs still challenge mobile hardware. Continued optimisation is a core research area.
Privacy Composition Challenges
Combining privacy protocols with DeFi is challenging. A user with shielded ETH in Railway who wants to use public DeFi can do so from a private 0zk address through Railway's DeFi integrations, but those interactions still run against public contracts, and exiting the pool reveals information; each transition adds cost. Native private DeFi (private swaps, private lending) requires application-level privacy support. This is a substantial multi-year effort. Aztec's aim is to enable this natively; Railway's aim is to provide privacy for existing token operations.
Regulatory and Business Uses
Beyond individual privacy, ZK privacy layers enable business use cases. Corporate treasury management without revealing positions. Institutional trading without public order flow leakage. Confidential procurement and supply chain management. These use cases have strong economic value but require production-grade infrastructure. Aztec and similar projects are positioning for this market. Business adoption is slower than retail but potentially larger and more sustainable.
Watching Privacy Layer Adoption
Post-Tornado privacy layer adoption is a leading indicator of DeFi privacy demand. If Aztec and Railway achieve substantial adoption (billions in TVL, meaningful transaction volume), it demonstrates real user demand for privacy. If they remain niche, it suggests privacy demand is lower than developers assumed. Watch adoption metrics (TVL, active users, transaction volume) closely over time to gauge the trajectory.
Key Takeaways
- Privacy has proving, gas, UX, and composability costs
- Client-side proving preserves privacy but requires user compute
- Composability with existing DeFi is limited without native privacy support
- Business/institutional privacy use cases may be larger than retail
- Adoption metrics will indicate real privacy demand
Related Content
Related Coins
Key Terms
More Topics
Browse all 189 topicsHalving Market Behaviour
How Bitcoin price has actually behaved around halvings (2012, 2016, 2020, 2024), the structural reasons behind the patterns, and what is known about the 2028 halving.
Monero Privacy — Deep Dive
How Monero's ring signatures, stealth addresses, and Bulletproof range proofs combine to provide default-on privacy for every transaction, and the regulatory pressures the project faces.
Zcash Architecture
How Zcash uses zk-SNARKs to enable optional shielded transactions, why Sapling was the breakthrough that made shielded pool usage practical, and how Halo2 removed the trusted setup.
Tornado Cash Mechanics and OFAC Context
How Tornado Cash's Merkle tree deposit-and-mix architecture worked, the August 2022 OFAC sanctions, the 2024 Van Loon ruling and the March 2025 delisting, and the broader question of sanctioning immutable smart contracts.
Lightning Network — Deep Dive
How Lightning's payment channels enable instant Bitcoin transactions, the routing challenges that shape the network's topology, and why LN adoption has grown more slowly than early forecasts suggested.
Liquid Network
How Liquid is a Bitcoin sidechain with faster settlement, confidential transactions, and asset issuance — and where it fits in the Bitcoin scaling ecosystem alongside Lightning.
Validium, Plasma, and State Channels
The scaling designs that lost to rollups — Validium (off-chain data), Plasma (mass exit games), and state channels (bilateral off-chain) — and why understanding them helps make sense of current rollup design.
App-Chains and Rollup-as-a-Service
How Cosmos-style app-chains and modern Rollup-as-a-Service (RaaS) providers (Caldera, Conduit, AltLayer, Gelato) let applications deploy dedicated chains, and the trade-offs vs shared chains.
References & further reading
- primary
- secondary
- primaryAztec - Roadmap
Ignition Chain Nov 2025, Alpha Network 31 Mar 2026, Tierra (Alpha V6) 3 Nov 2026, Beta V1 ~Q2 2027.
- primaryRailway Wallet - Private DeFi Wallet
Wallet app on the RAILGUN privacy system; zk-SNARK shielding into 0zk addresses; Private Proofs of Innocence.
- primaryUS Treasury - Tornado Cash Delisting (21 March 2025)
Removal of economic sanctions against Tornado Cash.