Vitalik Buterin backs Frame Transactions as Ethereum’s native AA debate intensifies

Ethereum’s developer community is currently debating on how to implement native account abstraction (AA) as planning for the upcoming Hegota hard fork continues.

Several proposals have emerged in recent weeks, with EIP-8141, known as “Frame Transactions,” gaining attention. The proposal has been formally introduced as a possible main feature of the upgrade.

Following the discussion, Vitalik Buterin publicly responded to the proposal, stating that Frame Transactions could support a wider range of privacy and censorship-resistant use cases while potentially simplifying wallet architecture.

Native account abstraction debate gains momentum

The push for native account abstraction has intensified over the past month, with multiple Ethereum Improvement Proposals introduced.

EIP-8141 proposes a model known as Frame Transactions, which differs from traditional transaction formats by removing embedded signature fields. Instead, signatures and authorization logic are passed as data to smart contracts that validate transactions.

The proposal introduces a new opcode, APPROVE, that enables smart contracts to authorize sending transactions, paying gas, or both.

In addition, this design enables transaction authorization to be processed by programmable logic rather than fixed transaction fields. According to the proposal, this structure is capable of supporting alternative signature systems, conditional gas sponsorship and privacy-focused transaction mechanisms.

— Derek Chiang | ZeroDev (@decentrek) March 9, 2026

For example, gas sponsorship could be arranged through contracts that pay network fees in return for token transfers, while authorization logic could be implemented using multisignature or alternative cryptographic schemes.

At the same time, the model presents new operational challenges. Due to the possibility of executing smart contract code during transaction validation, Ethereum clients would need additional protection against denial-of-service attacks from mempools.

Frame Transactions versus Tempo Transactions

The debate over native account abstraction implies two distinct design philosophies in Ethereum development.

One approach, represented by Tempo Transactions, is to embed commonly used account abstraction features directly into the protocol. These include gas abstraction, atomic batching multiple operations, transaction scheduling, and sponsored transaction fees.

Tempo-style transactions organize these features right into the transaction format. Fields like arrays of calls allow for atomic batching, and timestamp parameters can be used to support scheduled execution. Another signature field allows a third party to cover the gas costs by co-signing the transaction.

Developers promoting this model contend that having features built right into the protocol is simpler to integrate and enhances the user experience. However, the approach may be somewhat extensible because new features would require protocol upgrades.

Frame Transactions takes the opposite approach by having generalized primitives instead of predefined features. Authorization and gas payment logic can be implemented in smart contracts, allowing developers to create custom systems for signatures, permissions, and transaction validation.

Vitalik Buterin highlights privacy and wallet design implications

Responding to the ongoing discussion, Vitalik Buterin stated that Frame Transactions could also enable privacy-focused applications to operate without the need for public transaction broadcasters.

According to Buterin, the design enables privacy systems such as Railgun and other protocols to directly interact with network functionality, such as FOCIL, while preserving censorship resistance.

It’s a good post, and thanks for your contributions to improving frame txs!

I would also add:

* Frame txs are also meant to cover a long-tail of privacy and censorship resistance use cases. They enable Railgun, PP, etc to work without public broadcaster intermediaries, and…

— vitalik.eth (@VitalikButerin) March 9, 2026

He also identified potential changes to the wallet architecture. Buterin said the idea of “every wallet being a smart contract” has already been successfully implemented in other ecosystems, citing Bitcoin’s multisignature wallet design.

In his opinion, wallets built using EIP-8141 could be relatively simple and execute only a few operations, as Bitcoin scripts do.

Buterin said that many wallet features currently implemented in large smart contracts, such as transaction batching and signature hash calculations, could be moved out of wallet code using the proposed structure.