The "Trojan Horse" in Ethereum's Fusaka Upgrade: How to Turn Billions of Phones into Hardware Wallets?

Author: Zhixiong Pan

The hardware wallet is already in your pocket

Inside the smartphones and computers we use every day, there are dedicated security chips built in. For example, the "Secure Enclave" in iPhones, or Keystore / Trust Zone / StrongBox in Android devices.

This isolated physical area is commonly known as a TEE (Trusted Execution Environment). Its defining feature is "in-only, never out": private keys are generated inside and never leave this secure zone; externally, you can only request it to sign data.

This is effectively the standard of a hardware wallet. When signing, these chips typically use an industry-standard cryptographic curve selected by NIST (National Institute of Standards and Technology): secp256r1. This is also the foundation behind WebAuthn and FIDO2 (such as your fingerprint login or FaceID).

A chasm separated by just one letter

The awkward truth is that Ethereum natively does not support this mainstream curve, secp256r1.

Back when Bitcoin was developed, concerns about potential "nation-state backdoors" in NIST curves led the community to adopt the less common secp256k1. As a result, Ethereum followed suit and inherited this curve for its account system.

Although r1 and k1 differ by only one letter, mathematically they are entirely different languages. This creates a major pain point: the secure chip in your phone is essentially clueless when facing Ethereum—it cannot directly sign Ethereum transactions.

If we can't change the hardware, let's make this version compatible

Ethereum clearly can't force Apple or Samsung to redesign their chips to support secp256k1. The only viable path is for Ethereum itself to adapt to secp256r1.

Can we use smart contracts to verify r1 signatures? In theory, yes—but the mathematical operations are too complex. Running one verification could cost hundreds of thousands of Gas, making it economically unfeasible.

Thus, in the Fusaka upgrade, developers deployed a powerful solution: precompiles. This is like creating a "backdoor" or "plugin" within the Ethereum Virtual Machine (EVM). Instead of having the EVM compute step-by-step, the verification logic is implemented directly in the client’s底层 code. Developers simply call a specific address to perform the verification at extremely low cost.

In EIP-7951, this cost is fixed at 6,900 Gas—dropping from hundreds of thousands to just thousands—and finally entering the range usable in real-world products.

The final piece of account abstraction

The implementation of this EIP means we can now use the TEE environment on phones to sign and authorize transactions for Ethereum smart accounts.

It's important to note that this does not apply to your current MetaMask-style EOA addresses (since their public key generation still relies on k1).

This is specifically designed for "account abstraction" (AA wallets). In the future, your wallet won't be a seed phrase, but a smart contract containing rules such as:

"If this fingerprint (r1 signature) is verified, allow the transfer."

Summary

EIP-7951 may not eliminate seed phrases overnight, but it has finally removed one of the biggest roadblocks standing in the way of Ethereum’s mass adoption.

Before this, users were always faced with a harsh choice: Want "bank-grade" self-custody security? You must buy a OneKey, Keystone, or Ledger device and safeguard your seed phrase like gold bars. Want the smoothest experience? You have to keep your funds on exchanges or custodial wallets, sacrificing control (and decentralization).

After the Fusaka upgrade, this dilemma disappears.

With EIP-7951 live, "your phone is your hardware wallet" will gradually become reality. For the next billion new users, they might never need to know what a "private key" is, nor face the stress of writing down 12 words.

All they need to do is what they already do when buying coffee—scan their face or press their fingerprint. The iPhone's secure chip will then use secp256r1 to sign the transaction, and Ethereum’s native precompile will verify it.

This is how Ethereum should welcome the next billion users—not by arrogantly demanding they learn complex cryptography, but by humbly embracing internet-wide standards and stepping into their pockets.