Introduction: Why Ethereum Still Struggles With True Privacy

A few days ago, we talked about our collaboration with Fluton × Fhenix—a project exploring how confidentiality and anonymity can coexist in a composable DeFi system.

Here, I want to tackle the problem of anonymity more concretely: why it’s difficult, why it matters, and how we at Fhenix are approaching it.

What Anonymity Really Means

Anonymity means unlinkability—observers shouldn’t be able to tell which user sent or received funds (and, by extension, what assets they hold).

In traditional finance, that’s trivial: banks know your history, but the public doesn’t.

On blockchains, however, every transaction is public, deterministic, and permanently linked.

That’s the paradox of transparency: the same openness that gives us verifiability also destroys privacy.

UTXO Systems: Why Bitcoin and Zcash Have It Easier

In UTXO-based systems like Bitcoin or Zcash, anonymity comes almost for free.

There are no running account balances—just discrete “coins” representing unspent outputs.

When you spend a coin, you create a transaction consuming that output and producing new ones.

Example: if I have an unspent 10 BTC output and want to send 2 BTC, the transaction creates two new outputs—

• 2 BTC to the recipient
• 8 BTC back to myself as change

Because each coin is committed to a Merkle tree, zero-knowledge proofs (ZKPs) can efficiently prove correctness without revealing who sent what or how much.

That’s how Zcash achieves both anonymity and confidentiality at once—mathematical elegance born from the UTXO model.

The Account Model: Ethereum’s Beautiful but Leaky Design

Ethereum—and other account-based systems like Solana—take a very different approach.

Each address maintains a balance entry in the global state. Transactions simply decrease one account’s balance and increase another’s.

It’s clean, developer-friendly, and easy to reason about—but it’s stateful, and that’s the problem.

Because balances live in persistent state, transactions are inherently linkable. Even if amounts are encrypted using Fully Homomorphic Encryption (FHE) or partially homomorphic methods, the sender–receiver relationship remains visible on-chain.

How Current Privacy Tools Work—and Where They Break

Projects such as Tornado Cash, Railgun, and Privacy Pools attempt to re-create UTXO-style anonymity inside Ethereum.

Users deposit assets, then withdraw to new addresses, effectively simulating “notes.”

It works—but introduces painful trade-offs:

• Lost composability: Private tokens can’t interact natively with DeFi contracts.
• Broken programmability: You must exit privacy layers to use your assets.
• Limited cross-chain functionality: Bridges and protocols can’t easily interoperate.
• Clunky UX: Depositing and withdrawing breaks the smooth account-based flow.

In short, these systems achieve privacy by abandoning the very model that makes Ethereum powerful.

What We Actually Want: Native Account-Based Anonymity

What we really need is native anonymity—a privacy layer that works within Ethereum’s account model while keeping everything composable.

That means anonymous transactions that can still interact with DeFi, DAOs, NFTs, and cross-chain protocols.

It’s a much harder problem than UTXO anonymity, but solving it would unlock programmable privacy for the entire Ethereum ecosystem.

Research on the Frontier

Two recent academic works highlight both progress and challenges in this area:

1. High-Throughput Three-Party DPFs with Applications to ORAM and Digital Currencies

(My work with Avishai Wool and Alex Pentland)

This approach uses secure multi-party computation (MPC) to achieve extremely fast throughput.

However, MPC is limited by the number of servers involved—typically three—making it unsuitable for public, permissionless networks.

Read the paper →

2. PriFHEte: Achieving Full Privacy in Account-Based Cryptocurrencies Is Possible

An FHE-based feasibility result proving that full privacy can be achieved theoretically—though not yet practical at scale.

Read the paper →

These studies show that combining FHE with account-based anonymity is not only feasible, but within reach—given enough innovation.

From Theory to Practice: The Fluton × Fhenix Breakthrough

Together with Fluton, we’ve been testing how these ideas perform in practice.

Fluton’s architecture uses Fhenix’s coFHE technology to encrypt transaction details end-to-end, ensuring confidentiality, while smart-account abstraction provides unlinkable identities for each interaction.

This means users can trade, bridge, and interact across protocols while remaining invisible to outside observers—no mixers, no broken composability, and no privacy compromises.

This collaboration validated a core thesis:

Dual privacy—confidentiality + anonymity—can be achieved natively inside Ethereum’s account model.

It’s an important milestone, but also just the beginning.

How We’re Thinking About It at Fhenix

At Fhenix, we’re taking these findings further. Our work focuses on combining FHE, account abstraction, and intent-centric architectures to deliver native anonymity that’s practical, performant, and composable.

Our roadmap includes:

• Encrypted address computation for unlinkable transactions
• Privacy-preserving authentication using FHE + ZK hybrids
• Composability-first design so DeFi remains interoperable
• Encrypted RPC endpoints to eliminate metadata leaks

The end goal: programmable privacy infrastructure for the entire Ethereum ecosystem.

Conclusion: The Future of Programmable Privacy

Anonymity is the Achilles’ heel of Ethereum’s privacy stack.

But it’s also the most exciting frontier—because once solved, it completes the vision of fully private, composable DeFi.

At Fhenix, we believe privacy isn’t a feature. It’s infrastructure.

And native anonymity is the final piece of that infrastructure puzzle.

‍