The Ethereum upgrade EIP-7702 launched as part of the Pectra hard fork on May 7, 2025, marking a crucial moment in Ethereum’s journey toward account abstraction. Specifically, EIP-7702 gives smart contract account features to Externally Owned Accounts (EOAs). In the simplest terms, account abstraction aims to render Web3 so straightforward that users don’t even notice they’re using it, just like someone swipes their credit card to make a payment without understanding the transaction’s backend technicalities. The new protocol benefits developers by letting them integrate sophisticated functions directly into user accounts. These include wallet recovery mechanisms, automated transactions, etc.
Ambire Extension is the first wallet to support EIP-7702
Self-custodial wallet pioneer Ambire launched the browser-based Ambire Extension, a wallet with native EIP-7702 support, the day after Pectra. It does away with token approval, allows payment of gas fees in tokens like stablecoins, and brings smart account functionality to existing accounts, unlocking programmable, portable, and cost-efficient self-custody without needing to migrate accounts. Publicly available for Chrome without invite codes, the Ambire Extension combines the benefits of EOAs with smart account functionality, letting users keep existing addresses (reputations, whitelists, etc.) while unlocking programmable security, transaction batching, and gas abstraction. These powerful features were previously only available through smart contract wallets.
Essentially, users benefit from custom rules, automation, and security enhancements with the simplicity and familiarity of a regular wallet. Eliminating token approvals and allowing gas payments in other tokens reduces friction and potential user errors. Ambire is turning old wallets into next-gen smart wallets without any radical intervention from the user.
EIP-7702 combines the best features of previous standards
The ERC-4337 standard is associated with ownership and control decoupling risks and potential compatibility issues. It is dependent on the EntryPoint contract to process and validate UserOperations, which can become a point of failure. Account abstraction can decouple account control and ownership from EntryPoint, introducing smart contract bugs, unauthorized transactions, and other risks. Complexity rises as developers implement technical mechanisms to mitigate risks.
ERC-4337 offers complex functions, while existing EOAs are far simpler. This necessitates significant code refactoring and backend logic. Logistical challenges increase as developers need to decentralize bundlers for security and update MetaMask and other popular infrastructure. What’s more, smart contract transactions cost more than traditional EOA transactions.
EIP-3074’s AUTH and AUTHCALL opcodes enhance EOA capabilities, allowing them to act like smart contract accounts. The risk of a centralized point of failure still exists, as this protocol relies on invokers and requires a hard fork.
EIP-7702 overcomes the limitations of ERC-4337, EIP-3074, and EIP-5003 while combining their best features. It aims to improve the efficiency and compatibility of account abstraction implementation. The protocol’s delegation mechanism requires users to authorize smart contract code execution, and only they can revoke this authorization.
Execution starts with an EOA delegating control to a smart contract wallet. The system then validates or rejects the delegation, the EOA obtains smart contract capabilities, and another 7702 transaction is sent to set the delegation back to zero if the EOA decides to revoke smart contract control.
EIP-7702 is not without its critics. Traditional security checks relied on the assumption that EOAs were simple, codeless wallets. Now that they can respond to calls within transactions, checks that assume they cannot run logic become unreliable. For instance, extcodesize checks if an address is a contract, but EOAs appear as contracts with code, and measures based exclusively on code size can yield false negatives. Many DeFi protocols use tx.origin to prevent reentrancy attacks, assuming these are always initiated via contracts. After EIP-7702, attackers can initiate transactions that only seem to come from an EOA and execute malicious reentrant calls.
Pectra is the most feature-packed upgrade to date
Pectra combines the Electra consensus layer upgrade and the Prague execution layer hard fork. It’s the first upgrade since Dencun in March last year and Ethereum’s most feature-packed one in history: 7702 is just one of 11 included EIPs. Their features will drive the network forward and improve staking, making it much easier for institutional and retail investors to secure Ethereum. Improved integration with L2 blockchains promises to transform network efficiency.
The second EIP is EIP-7002, which enables smarter, programmable staking by letting validators exit through the execution layer. They were limited to consensus-side mechanisms before. The new flexibility makes it possible to develop automated strategies, smart contract-based staking managers, and other advanced use cases.
EIP-7251 raises the staking limit from 32 ETH to 2,048 ETH, improving efficiency by reducing the validator count. The 32 ETH cap caused the number of validators to skyrocket, raising hardware requirements for node operators and stressing the network.
EIP-6110 simplifies onboarding by moving validator deposits to the execution layer. In the past, they were relayed through the consensus layer, leading to delays and unnecessary complexity.
EIP-7840 supports scaling upgrades by standardizing blob scheduling across clients. Clients had no reliable way of scheduling blob changes before, and coordinating blob capacity upgrades was a hassle. EIP-7691 doubles the network’s target blob count per block (from 3 to 6) and raises the upper limit to 9, increasing the space available for L2 rollups to post data. Before this change, rising rollup demand caused blob space bottlenecks. It also helps improve scalability and reduce L2 fees.
EIP-7623 is another upgrade associated with better scalability. It raises call data fees to encourage the use of blobs for rollup data. Arbitrum, Optimism, and other L2 solutions posted data using call data in the past, which was costlier and less efficient than using blobs. Users now benefit from lower L2 fees.
EIP-7549 moves the committee index out of attestation, improving consensus efficiency. Before the change, attestations were harder to optimize due to their excessive size. The network carried redundant data in its consensus messages, adding to storage and bandwidth requirements.
Before EIP-7685, communication between the execution and the consensus layer was unstandardized and limited. With this proposal, future upgrades relying on more solid interlayer data sharing will be easier to implement.
EIP-2537 solves the challenge of efficient BLS signature verification. Previously, cryptographic operations were too gas-intensive for widespread adoption.
Ethereum limited smart contracts to the 256 most recent block hashes before Pectra’s final proposal, EIP-2935, making it harder to build apps relying on older data. The upgrade opens up use cases in trustless oracles and proof systems.
