Ethereum researchers are working to resolve a critical cryptographic proof gap in zkEVM security, aiming to meet the 128-bit system-level requirement before the December 2026 deadline. The outcome could shape the future of Ethereum scaling and security.
Ethereum's push to deliver a production-ready zkEVM with 128-bit provable security is facing a critical test as researchers race to close a persistent cryptographic proof gap before the December 2026 target. The better.codes contest, launched by the Ethereum Foundation, now provides a public, reproducible benchmark for measuring the gap between what has been formally proven safe and what remains potentially vulnerable in the system's core proof architecture.
Proof Gap and Security Stakes
As of August 21, the live leaderboard for the koalaIRS12 benchmark-a fixed parameter profile for an interleaved Reed-Solomon reduction-showed a 63.99-bit lower certificate and a 116.13-bit upper certificate. This leaves a 52.14-bit interval where neither side has been definitively proven, far short of the 128-bit system-level security target set by the Foundation. The contest's two-track approach allows researchers to either raise the lower bound by proving more of the system is safe (soundness track) or lower the upper bound by identifying unsafe regions (attack track). Each submission is machine-verified using the Lean kernel, ensuring that only rigorously checked results are promoted to the leaderboard.
Technical and Practical Hurdles
The Ethereum Foundation has emphasized that closing this gap is not just about hitting a numerical target. Each certificate applies only to the specific parameter point encoded in the challenge, and broader system security will require additional analysis across all components, including proof size, recursion architecture, and implementation fidelity. The Foundation's roadmap, updated in February, moved the M3 deadline to early December 2026 and requires teams to connect component-level proofs to an auditable system package. For koalaIRS12, reaching the 128-bit lower certificate would settle the benchmark at its fixed parameter, but a production zkEVM claim will demand comprehensive evidence that all components meet the required standard.
Research, Review, and Open Questions
Recent academic work by Gal Arnon, Dan Boneh, and Giacomo Fenzi has highlighted unresolved questions in succinct proof systems, including list decoding and Reed-Solomon proximity gaps. The Ethereum Foundation's own review of the SP1 formal-verification effort in May underscored the need for reproducible pinning of code, inputs, and versions, as well as broader reasoning to support system-level conclusions. The better.codes contest is framed as a machine-checked research path for hash-based SNARK security, with each leaderboard movement providing new evidence for future security arguments.
Roadmap, Progress, and Industry Context
The Foundation's public progress page, last updated August 20, tracks zkVM readiness and integration milestones but does not yet mark completion for the full early-December package. Optional execution proofs are currently supplementing mainnet testing, while consensus-critical validation remains unchanged. The 52.14-bit unresolved interval on koalaIRS12 is now a live measure of unfinished work, and the outcome will influence whether Ethereum's zkEVM can meet its 128-bit production security goal. This technical race comes as other major networks are also re-evaluating their security and scaling strategies-GnosisDAO, for example, recently approved a plan to sunset its Layer 1 blockchain and shift to an Ethereum rollup, raising new questions about network security and token utility (see EgonCoin's coverage of the GnosisDAO transition).
According to the latest available data, the better.codes leaderboard for koalaIRS12 as of August 21, 2026, showed a 63.99-bit lower certificate and a 116.13-bit upper certificate, with nine promoted submissions from seven solvers. The Ethereum Foundation's roadmap calls for a 128-bit provable security threshold, a final proof size of 300 KiB or less, and a formal soundness argument for the recursion architecture, all to be delivered by December 1, 2026.
Cryptographic proof systems like those used in zkEVMs rely on formal reductions to establish security guarantees. However, these reductions are only as strong as their weakest component, and gaps between lower and upper certificates represent areas where the system's safety cannot be fully assured. Closing these gaps requires not only mathematical rigor but also careful attention to implementation details, parameter choices, and the composition of independently analyzed modules. As Ethereum and other networks pursue scalable, secure zero-knowledge solutions, the ability to provide machine-checked, system-level proofs will remain a central challenge for both researchers and developers.
