Introducing Encrypted Transactions

Blockchains made financial systems programmable, global, and always-on. They also made them radically transparent. For institutions, this creates a structural mismatch. Every balance, every transfer, every flow is publicly visible by default. Treasury movements can be tracked. Trading activity can be inferred. Payment relationships can be mapped. This is not how financial systems are designed to operate, and this is not how regulators expect the world to run in the near future.

Today, we are introducing Encrypted Transactions in Dfns. Starting with Ethereum and EVM-compatible networks, Dfns now enables clients to send transactions where amounts and balances are encrypted onchain, using the emerging ERC-7984 confidential token standard and Zama’s Fully Homomorphic Encryption (FHE) technology.

From transparency to programmable privacy

Public blockchains expose state in plaintext. ERC-7984 changes that by introducing a standard for confidential tokens, where balances and transfer amounts are represented as encrypted values instead of integers. These encrypted values are not just hidden, they remain usable.

Smart contracts can still validate transfers, enforce rules, and update balances without ever decrypting the underlying data. This is a fundamental shift. It means privacy is not achieved by moving off-chain or reducing functionality, but by changing how computation itself is performed. Instead of reading balance = 100, contracts manipulate encrypted handles that represent that value, and correctness is enforced cryptographically.

At the core of ERC-7984 is Fully Homomorphic Encryption (FHE), implemented by Zama. FHE allows computation directly on encrypted data. In practical terms, this means that addition, subtraction, and other operations can be performed on ciphertexts, producing encrypted results that, once decrypted, match what would have happened in plaintext.

This is fundamentally different from traditional encryption models, where data must be decrypted before it can be used. With FHE, the data never leaves its encrypted form during computation. In the context of ERC-7984, this enables:

• encrypted balances stored on-chain
• encrypted transfer amounts
• smart contract execution over ciphertexts
• verifiable state transitions without revealing underlying values

The standard defines how encrypted values are referenced for FHE computation, how they are passed between contracts, and how operations are performed consistently across implementations.

Tooling from OpenZeppelin extends this model with reusable contracts and patterns, including wrappers and permission systems for accessing encrypted data.

How encrypted transactions work in Dfns

Dfns integrates ERC-7984 directly into its wallet and transaction stack. From a developer perspective, the workflow remains familiar. You still create wallets, construct transactions, sign them, and broadcast them. The difference lies in how the payload is built and interpreted. When interacting with confidential tokens:

• transaction amounts are encrypted before submission using Zama relayers.
• balances are stored onchain as encrypted values (“FHE handles”)
• contracts process encrypted inputs and produce encrypted outputs

Dfns handles the blockchain-specific complexity required to make this work in production. This includes transaction serialization, encoding of encrypted parameters, signature normalization, and compatibility across EVM chains.

A key concept in ERC-7984 is read permissions. Because balances are encrypted, reading them requires explicit authorization. Wallets can delegate read access to specific actors, including Dfns (see code examples here), allowing decrypted balances to be retrieved when needed for operations or UI display.

If no read permission is granted, Dfns stores and tracks the encrypted value directly, ensuring that transaction integrity and accounting remain consistent even without access to plaintext data. This allows teams to choose their privacy model, from fully opaque to selectively readable systems, without changing infrastructure.

A practical model for institutional privacy

Encrypted transactions enable a range of real-world use cases that were previously incompatible with public chains. Treasury operations can be executed without exposing allocation strategies. Trading activity can occur without leaking intent or position sizing. Payment flows can remain confidential while still settling on public infrastructure. Because ERC-7984 is an EVM standard, this model applies across:

• Ethereum mainnet
• Layer 2 networks such as Optimism, Arbitrum, and Base
• EVM-compatible private networks such as Hyperledger Besu

This ensures that encrypted transactions are not tied to a single environment, but can be deployed across the full spectrum of EVM-based infrastructure.

We are starting with Zama’s FHE implementation because it provides the most complete, production-proven, and standardized approach for encrypted computation on EVM today. That said, encrypted transactions are not limited to a single cryptographic technique. Dfns is building support for additional privacy primitives, including other zero-knowledge and encryption-based approaches. Each of these brings different trade-offs in terms of performance, expressiveness, and trust assumptions.

Our goal is not to enforce a single model, but to provide a unified interface across multiple privacy technologies, allowing institutions to choose what fits their requirements without changing how they build or operate their systems.

Current limitations and trade-offs

Though Zama is running and maintaining the code to ensure security and efficiency, FHE operations are computationally intensive, which can introduce performance overhead compared to standard transactions. While this is improving rapidly, it remains a factor in system design.

The ecosystem is also still maturing. ERC-7984 is an emerging standard, and tooling, SDKs, and contract patterns are still evolving. This means developers need to understand new primitives such as encrypted handles, permission delegation, and ciphertext-based logic.

These limitations are known and actively being addressed. Improvements in FHE performance, standardization, and developer tooling are expected over the coming months, and Zama and Dfns are contributing to making these systems more accessible and production-ready.

A new default for onchain finance

Encrypted transactions redefine what it means to use public blockchains. They preserve the core properties that make blockchains valuable: programmability, composability, and verifiability. At the same time, they remove one of their most limiting characteristics: full transparency by default.

Access to data becomes a controlled capability rather than a default. Because balances are encrypted, visibility is granted through explicit access controls. This is not a limitation but an improvement as it allows institutions to define precisely who can see what, aligning privacy with compliance requirements. Though it does require adapting to a new paradigm where data access is intentional, auditable, and policy-driven, we believe that this is the direction onchain finance should take to achieve mass adoption.

At Dfns, we see this as a foundational evolution of wallet infrastructure. Encrypted transactions are now available on Dfns for Ethereum and EVM-compatible networks.

• Get started today: app.dfns.io
• ‍Learn more about Zama and FHE: zama.org