Vitalik Buterin Says zkEVM Security Must Take Priority After ZKsync Airbender Milestone
Ethereum co-founder Vitalik Buterin has called for security to become the main priority for zero-knowledge Ethereum Virtual Machines (zkEVMs), congratulating the ZKsync Airbender team on a recent milestone. In his statement, Buterin argued that the competition to improve zkEVM speed has reached a decisive stage and that development efforts should now concentrate on proving systems secure.
Buterin outlined three areas that he believes require formal verification: equivalence between a zkEVM’s arithmetization and the Ethereum specification, a strict upper bound on prover costs relative to the gas processed in a block, and the security of the proof system itself. His comments focus on the mathematical and technical foundations needed to ensure that faster systems remain reliable.
The statement highlights a central challenge for Ethereum scaling technology. Faster proof generation can improve the practicality of zero-knowledge systems, but speed alone does not guarantee that execution is correct or that cryptographic proofs are secure. Buterin’s message suggests that the next phase of zkEVM development should place greater emphasis on independently verifiable guarantees.
Vitalik Identifies Three Priorities for zkEVM Security
The first priority identified by Buterin is formally verifying that a zkEVM’s arithmetization is equivalent to the Ethereum specification. Arithmetization translates program execution into mathematical constraints that a proof system can check. If those constraints do not accurately represent Ethereum’s execution rules, a system could potentially prove that an incorrect computation satisfies its own model.
This issue matters because zkEVMs aim to reproduce Ethereum-compatible execution while using cryptographic proofs to verify computations efficiently. Formal verification can help establish that the mathematical representation matches the intended specification, reducing the risk of errors between what Ethereum requires and what a proving system actually checks.
Buterin’s second priority is establishing a hard maximum bound on the ratio of prover cost to the amount of gas in a block. Generating zero-knowledge proofs requires computational resources, and those costs affect whether a system can operate reliably at scale. A defined upper bound could help developers assess the relationship between the work represented by a block and the resources required to prove its execution.
The third priority is verifying the security of the proof system itself. Even when execution constraints accurately represent Ethereum and proving costs remain manageable, the underlying cryptographic system must still resist attacks that could allow invalid statements to be accepted as valid. This makes proof-system security a separate concern from execution equivalence and performance.
Taken together, the three requirements address different layers of risk: correctness of the execution model, predictability of proving costs and the integrity of the cryptographic proofs. Buterin’s position is that these properties should be formally established rather than treated as secondary considerations after performance targets have been met.
ZKsync Airbender Milestone Puts the Security Debate in Focus
Buterin directed his congratulations to the ZKsync Airbender team but did not provide technical details about the milestone in the statement. His message also did not claim that Airbender had already satisfied all three proposed security requirements. The comments instead set out what he believes should become the broader priority for zkEVM development.
The distinction is important because faster proof generation and stronger security guarantees are related but separate achievements. A system may improve proving speed without having completed formal verification of every relevant component. Similarly, formal verification of one layer does not automatically establish the security of the entire system.
For Ethereum developers and scaling projects, this approach could mean greater attention to rigorous specification matching, transparent cost limits and scrutiny of cryptographic assumptions. Such work can be demanding, but it addresses failure modes that performance benchmarks alone cannot reveal.
The implications also extend to users and applications relying on zkEVM infrastructure. Security weaknesses in execution or proof verification can undermine the assurances that these systems are intended to provide. Stronger verification methods could help improve confidence, although no single formal verification effort can eliminate every implementation, operational or cryptographic risk.
Buterin’s declaration that the speed battle has been won represents his assessment, not a universally established conclusion across every zkEVM design. His central argument is clearer: as performance improves, the industry should not treat speed as the final measure of progress. The next milestone should be demonstrating that zkEVMs execute Ethereum-compatible computations correctly, operate within defensible proving-cost limits and rely on secure proof systems.
About The Author
