Ethereum Foundation researcher Justin Drake said Ethereum is accelerating its transition to post-quantum cryptography, with developers aiming to complete the work by December 2028. The initiative is intended to protect Ethereum’s execution, consensus and data layers as advances in quantum computing and artificial intelligence create new risks for the network’s cryptographic systems.

In a presentation, Drake said AI-assisted research had substantially increased the number of vulnerabilities identified in recent months. Since March, developers of Ethereum clients have received 895 potential issues, of which 500 were confirmed as genuine vulnerabilities.

Drake said Ethereum benefits from having multiple clients. He added that the Ethereum Foundation’s security team had expanded in recent weeks and months and recruited researchers who specialize in AI.

Developers assess quantum-computing risks

Drake said the estimated timeline for quantum computers capable of breaking modern cryptography could shorten. He cited the head of a major quantum company who, according to Drake, expressed 99% confidence that “Q-Day” — when quantum computers can break the relevant cryptographic mechanisms — would arrive as early as 2029.

Drake placed his own baseline estimate at 2032 and assigned a 10% probability to Q-Day arriving by 2029. He also identified neutral-atom quantum systems as one of the main short-term cryptographic threats because of their scaling potential.

Progress is occurring across quantum hardware, error correction and algorithms, Drake said. He pointed to work on Shor’s algorithm, which can be used to break elliptic-curve cryptography.

According to Drake, about 100 enthusiasts working with the ecdsa.fail project reduced the estimated number of required qubits by 30% compared with a Google study published in March. He said they also cut the algorithm’s runtime threefold, while newer work mentioned in the presentation could reduce it by another order of magnitude, from minutes to seconds.

Drake separately warned that mathematical superintelligence could find counterexamples to conjectures long considered correct, raising further questions about the reliability of Ethereum’s cryptographic mechanisms. He also suggested that government bodies could restrict publication of research into cryptographic vulnerabilities. Both points represented Drake’s assessments.

Mission 2028 sets three-layer transition

Developers now want to protect all three Ethereum layers before the JAR fork. They previously planned the relevant changes for the later LAR fork.

The plan calls for post-quantum transactions at the execution layer, post-quantum attestations at the consensus layer and post-quantum data sampling at the data layer. LeanVM will serve as the primary implementation tool and, according to the presenters, recently achieved full end-to-end formal verification.

Mission 2028 includes two principal milestones: launching the Glamsterdam fork on mainnet in January 2027 and completing the transition to post-quantum and post-AI cryptography by December 2028. Developers plan to deploy the changes through three hard forks during the 24 months after Glamsterdam.

Hash-based systems and formal verification

A second speaker, Kev, said hash-based cryptography and formal verification would underpin the transition. The team considered Poseidon2 for its performance but moved toward classical hash functions as proof systems developed. Its primary current option is SHA-3/Keccak, which has undergone more than 10 years of cryptanalysis.

Developers plan to replace ECDSA with leanSphincs at the execution layer, BLS with leanXMSS at the consensus layer and KZG with leanDA at the data layer. They are also developing leanPCS, a post-quantum polynomial commitment scheme intended to support a move from Groth16 and Plonk to post-quantum STARK proofs.

Steps for validators, applications and users

Developers urged validators to check legacy 0x00 withdrawal credentials, which the speakers said are scheduled for deprecation, and to participate in the PQ Validator Registry. Application developers should review their systems for hard dependencies on ECDSA, while wallet developers should assess support for Frame Transactions, which are expected to become a central part of post-quantum protection at the execution layer.

Drake estimated that accounts which have signed at least one transaction hold 72% of ETH, meaning their public keys have been exposed. He suggested that users seeking to reduce this risk immediately could move their ETH to a new account that has never signed a transaction. The blockchain would then expose the address hash rather than the raw public key.

Users could alternatively use Frame Transactions with a smart contract that verifies leanSphincs signatures. According to the speakers, this approach can provide post-quantum protection now, although it will remain more expensive until leanVM is integrated directly into the protocol.

Developers also asked the community to provide feedback on leanSphincs parameters and test the system on the Daugi network.

Stacks founder Muneeb Ali previously said the quantum threat was already limiting bitcoin’s growth.

Source: Incrypted