EIP-7702, which shipped with Ethereum’s Pectra upgrade earlier this year, lets externally owned accounts temporarily behave like smart contracts during a transaction. That might sound like plumbing. It is plumbing - but the kind that changes what you can build on top.

Before Pectra, wallet UX on Ethereum was constrained by a simple fact: regular user wallets couldn’t execute logic. They could sign and send, but any conditional behavior - batching transactions, setting spending limits, automating gas payment in tokens other than ETH - required either a separate smart contract wallet like Safe, or a layered abstraction system. EIP-4337 tried to solve this at the application layer without a protocol change, and it worked, but it added complexity and relayer infrastructure that smaller teams found difficult to ship around.

7702 sidesteps that by letting a standard EOA delegate to a contract implementation for the duration of a transaction. Users don’t need to migrate to a new wallet type. Their existing address works. The wallet software handles the delegation behind the scenes.

What This Actually Enables

The most immediate use case is transaction batching. Approving a token and then swapping it in a single user confirmation - rather than two separate transactions - has been a persistent friction point in DeFi. With 7702-compatible wallets, that becomes straightforward without requiring users to move funds into a smart contract wallet first.

Sponsored gas is the other obvious unlock. Applications can pay transaction fees on behalf of users, removing ETH-in-wallet as a prerequisite for onboarding. This matters most at the edges of the ecosystem - gaming applications, consumer apps, any context where asking someone to acquire ETH before doing anything is a conversion killer.

The Wallet Side Is Slower to Move

Protocol support is live, but wallet adoption is uneven. MetaMask has shipped initial 7702 support. Rabby has moved in this direction. Hardware wallet compatibility - where signing delegation transactions requires firmware updates - remains an open coordination problem that hasn’t been fully resolved as of this writing.

The pattern here is familiar. Ethereum ships a capability at the protocol level, then waits for the tooling layer to catch up. That gap between “technically possible” and “usable by default” is where most of the real work happens, and it tends to happen slowly and without much announcement.

Account abstraction has been an Ethereum goal since at least 2018. 7702 is the first version that doesn’t ask users to change their address to get it.