Vitalik: AI will become the new user interface, directly executing complex on-chain operations or becoming the norm

At the OKX Now 2026 Summit in Singapore, Ethereum co-founder Vitalik Buterin stated that AI is demonstrating the ability to break out of sandboxes, attack websites, and discover software vulnerabilities, while also helping the Ethereum protocol identify vulnerabilities, implement complex cryptography, and conduct formal verification.

He believes that higher levels of security will become feasible and necessary, as AI will continuously discover exploitable vulnerabilities within systems.

Vitalik revealed that about a month ago, he updated ENS information for the first time without using a user interface (UI), only requiring a local AI agent to write a script, which was completed in about 5 minutes. He anticipates that executing complex on-chain operations directly through AI will gradually become the norm, with AI becoming a new user interface in more scenarios.

He cited the Safe-related hacking incident involving approximately $1.4 billion in losses as an example, pointing out that the attacks targeted the interaction layer between users and on-chain systems, rather than the on-chain contracts. Directly handling transactions with AI may reduce the risks associated with reliance on web interfaces, but it will also introduce new challenges, including whether the AI itself is secure, whether it can resist prompt injection, and whether it correctly understands on-chain operations. As interaction methods change rapidly, some issues will be resolved, while new security challenges will also emerge.