Vitalik Buterin Says AI Is Speeding Up Ethereum Development

• Agentic coding produced a 700,000-line Ethereum client prototype aligned with the 2030 roadmap in weeks.
• Buterin urges pairing faster AI coding with formal verification, expanded test suites, and multi-client checks.
• He also detailed a quantum-resistance roadmap replacing BLS and adding hash-based and STARK-based systems.

Ethereum co-founder Vitalik Buterin said artificial intelligence is fast accelerating Ethereum development. In recent public remarks, he described how agentic coding enabled a developer to build a roadmap-aligned Ethereum client within weeks. The work took place online in early 2026 and focused on speeding development while improving security through testing and verification.

AI-Driven Coding and Testing Gains

According to Buterin, a developer used agentic coding to produce a prototype Ethereum client aligned with the 2030 roadmap. Notably, the prototype included roughly 700,000 lines of code and covered 65 roadmap items. It also synced with Ethereum mainnet within two weeks.

However, Buterin stressed the build came with major limitations. He said the code likely contained critical bugs and incomplete implementations. Still, he explained that such progress seemed impossible six months earlier, highlighting how quickly development tools are advancing.

Buterin added that AI accelerates coding across tasks. He cited his own experience recreating blog software within an hour using local models. Therefore, he argued that speed gains should split evenly between faster development and stronger security practices.

Formal Verification and Security Focus

Buterin said the most effective AI use combines faster coding with deeper testing. Specifically, he highlighted generating larger test suites and expanding formal verification. He noted that a Lean Ethereum collaborator used AI to create a machine-verifiable proof for a complex STARK-related theorem.

Lean Ethereum, according to Buterin, aims to formally verify every component. AI, however, now accelerates that effort significantly. In addition, he emphasized that bug-free code remains difficult, but more achievable through repeated testing and multi-implementation checks.

He cautioned that secure systems will not emerge from single prompts. Even so, he said debugging cycles can now happen faster and more thoroughly.

Quantum Resistance Roadmap Details

Separately, Buterin outlined Ethereum’s quantum resistance roadmap. He identified vulnerabilities in BLS consensus signatures, KZG-based data availability, ECDSA account signatures, and zero-knowledge proof systems. To address these, he proposed replacing BLS with hash-based signatures and using STARK aggregation.

For externally owned accounts, Buterin pointed to EIP-8141. This change would enable multiple signature schemes, including quantum-resistant options. He acknowledged higher gas costs but said efficiency should improve over time.

The roadmap also addresses proof systems. While STARK proofs cost more than current ZK-SNARKs, aggregation could reduce on-chain computation. Last month, the Ethereum Foundation said the next ecosystem phase will expand capacity while maintaining long-term security and resilience.