Ethereum launches AI platform to advance post-quantum security

The Ethereum Foundation has launched better. codes, a collaborative research platform that utilizes artificial intelligence to advance the security of hash-based SNARKs. Developed in partnership with Yukon and zkSecurity, the initiative is part of Ethereum’s broader roadmap to transition toward post-quantum cryptography.

The platform focuses on the koalaIRS12 Reed–Solomon proximity problem, which is critical to modern cryptographic proof systems. Researchers can deploy AI agents to refine the soundness bounds of koalaIRS12 proofs, with all results undergoing machine-verified security checks via the Lean kernel. The project aims to establish a 128-bit security benchmark.

This shift is a proactive measure against future quantum computing threats. Current blockchain infrastructure relies heavily on elliptic-curve-based cryptographic primitives for transaction signatures and consensus mechanisms. These systems could potentially be compromised by Shor’s algorithm if implemented on sufficiently powerful quantum computers. By prioritizing hash-based alternatives like SNARKs and STARKs, Ethereum aims to mitigate these long-term risks.