Ethereum Weighs Aggregating Signatures and Proofs as It Prepares for Quantum Computing Risks
Key Facts
Discussion around leanPCS has highlighted EIP-8288, a proposal to aggregate signatures and proofs in Ethereum as the network prepares for quantum computing risks. The draft says block validity would require a recursive STARK proof in the block header demonstrating the validity of all transaction dependencies. According to the proposal, this design would allow many thousands of signatures and STARK proofs to be processed in each slot without excessive data overhead. But the proposal does not establish that this mechanism has been implemented, nor does it confirm that Buterin made a statement supporting it on November 16. Activating EIP-8288 would require a network upgrade, because nodes that do not implement the change would reject blocks containing the recursive-proof field.
The case for aggregation stems from the size of the signatures and proofs under discussion. According to EIP-8288's rationale, current proposals for hash-based quantum-resistant signatures range from 2 to 3 kilobytes and require between 150,000 and 200,000 gas units to verify. STARK proofs exceed 128 kilobytes and can reach 512 kilobytes when generation speed is a priority. The proposal says its mempool transaction aggregation design sets a fixed STARK-related burden of about 256 kilobytes per aggregation period for the block builder. These figures illustrate the technical trade-off the proposal seeks to address, but do not mean user transaction fees will fall.
The Ethereum Foundation describes the protocol transition as long-term, proactive work, rather than a response to imminent quantum computing capability. The foundation's post-quantum team said: "We don't believe a cryptographically relevant quantum computer is imminent, but migrating a global, decentralized protocol takes years." ethereum.org says current quantum devices are far from the level needed to break Ethereum's cryptography. Still, a study by Google Quantum AI published in March 2026 estimated that breaking 256-bit elliptic-curve cryptography could require about 1,200 logical qubits. The page says the U.S. National Institute of Standards and Technology expects ECDSA to be phased out by 2030 and prohibited by 2035.
The Ethereum Foundation aims to make the layer 1 quantum-resistant across the execution, consensus and data layers by December 2029, according to protocol priorities published on September 7. The plan assumes a "quantum day" could come as early as 2030, an assumption the foundation described as deliberately aggressive. The post-quantum team says layer 1 upgrades could be completed by 2029, while full migration of the execution layer would take additional years. The foundation says client teams may begin implementing Hegotá in late Q4 2026; the schedule, from the Glamsterdam release in December 2026 to L* in December 2029, implies an average pace of 7.2 months per fork. The roadmap lists post-quantum key registries at I*, precompile signature verification at J*, and full aggregation and quantum-safe data commitments at M*, while cautioning that planned dates may change.
Signature size presents another challenge in the transition to quantum-resistant alternatives. According to ethereum.org's comparison, hash-based signatures for validators are about 3,000 bytes, compared with 96 bytes for BLS signatures. The page describes leanVM as an aggregation engine that compresses data by a factor of 250. In a February 2026 protocol priorities update, the foundation linked native account abstraction with the move away from ECDSA-based authentication and cited proposals under development to improve the efficiency of verifying quantum-resistant signatures inside the Ethereum Virtual Machine. This is development work within a broader path; EIP-8288 itself does not guarantee lower fees or improved scalability.
EIP-8288 was created on June 3, 2026, as a draft core proposal co-authored by Vitalik Buterin and Tomasz Korda. The proposal links reuse of Lean Ethereum tools with handling large numbers of signatures and proofs while limiting data growth. Its adoption on the network remains contingent on developer consensus and a consensus upgrade, as unupdated nodes would reject blocks containing the new field. The foundation targets completion of layer 1 upgrades by 2029, while completing the execution-layer transition will take longer. The proposal therefore remains part of a broader engineering path to prepare for a post-quantum future, rather than an active change to the network.