Vitalik Buterin Updates Ethereum Roadmap With Quantum Safety, Privacy Focus

MakeCryippreferred on

Ethereum co-founder Vitalik Buterin published an updated version of his 2023 protocol roadmap on Aug. 10, overlaying it onto the Ethereum Foundation’s newer draft roadmap, the Strawmap, and highlighting where priorities have shifted. Quantum safety and privacy protections moved up the list, while native rollups, aggressive post-quantum scaling, and AI-assisted formal verification appear as additions that didn’t exist in the original plan.

Buterin posted the comparison on X, saying the overlap between the two roadmaps is substantial but that “some things got reshuffled in order,” pointing to quantum safety specifically, while other 2023-era items, including verifiable delay functions and a number of planned EVM improvements, were dropped.

Quantum safety moves to the front

The clearest shift is quantum resistance. Buterin authored EIP-8288 on June 5, a proposal that adds a new transaction frame type letting signatures and STARK proofs be aggregated into a single recursive proof rather than verified individually on-chain, which he refers to as “zkzk frames.” The goal is to make post-quantum signature schemes, which tend to be far larger and more expensive to verify than Ethereum’s current elliptic-curve signatures, cheap enough to use at scale.

Separately, Nicolas Consigny, who leads the Foundation’s Kohaku project, said in June that Ethereum could add post-quantum protection to individual accounts today, without a hard fork, for about 7 cents per account using a signature scheme called SPHINCS-. Buterin’s roadmap points to a longer-term successor, leanSPHINCS, meant to bring that cost down further through aggregation across many accounts at once. The Ethereum Foundation has cut 54 jobs, reducing its workforce by about 20% as the nonprofit restructures to streamline protocol development and ecosystem support.

Privacy becomes a first-class goal

Buterin’s update lists keyed nonces and recent roots, pieces of fork-choice enforced inclusion lists (FOCIL, defined in EIP-7805), and a mechanism he calls “lean privacy pool and wormholes.”

The wormholes concept traces to EIP-7503, first proposed in August 2023 by a team including Keyvan Kambakhsh and Hamid Bateni of Nobitex Labs. It lets a user send ether to an address with no known private key, effectively destroying it, then later prove with a zero-knowledge proof that they burned funds somewhere without revealing which address, and mint new ether at a different address instead. Because the system defaults every dormant Ethereum account into the same anonymity pool, there’s no way to prove any specific account participated.

Native rollups and an open question about the EVM

Buterin’s update also names native rollups, an approach where a rollup’s execution is verified directly by Ethereum’s own validators rather than through a separate bridge contract. He had argued as recently as February that the earlier model, where a high-throughput layer-2 network is secured only by a multisig bridge back to Ethereum, “no longer makes sense” if it isn’t backed by Ethereum’s own validator set.

That native-rollup direction feeds into a larger, still-unresolved question Buterin raised: whether Ethereum’s execution layer should eventually expose a simpler base instruction set, either leanISA or RISC-V, to users directly, with the EVM demoted to a compatibility layer running on top. No EIP has been filed for this yet, and Buterin described the deeper version of this idea as too early even for the Strawmap’s own draft timeline.

Why “lean” is doing more work than it sounds like

Buterin also flagged what he calls “lean-ification of the spec,” simplifying Ethereum’s protocol specification to make it possible to formally verify, mathematically prove correct, far more of the system than before. He credited this to advances in AI-assisted formal verification tools.

The idea traces to a July 2025 Ethereum Foundation post by researcher Justin Drake, “lean Ethereum,” which framed the consensus, data, and execution layers around minimalism and provable security rather than raw feature growth.