Vitalik: AI will become the new user interface, and directly executing complex on-chain operations may become the norm.

Ethereum co-founder Vitalik Buterin stated at the OKX Now 2026 Summit in Singapore that AI is demonstrating capabilities to break out of sandboxes, attack websites, and uncover software vulnerabilities, while also assisting the Ethereum protocol in identifying vulnerabilities, implementing complex cryptography, and conducting formal verification. He noted that a higher level of security will become both feasible and necessary, as AI will continuously discover exploitable vulnerabilities in systems. Vitalik revealed that about a month ago, he updated ENS information for the first time without using a user interface (UI), by only instructing a local AI agent to write a script, completing the operation in roughly five minutes. He anticipates that directly executing complex on-chain operations via AI will gradually become the norm, with AI emerging as a new user interface across more scenarios. Citing the Safe-related hack that caused around $1.4 billion in losses as an example, he pointed out that the attack targeted the interaction layer between users and on-chain systems, not the on-chain contracts themselves. AI directly handling transactions may reduce risks tied to reliance on web interfaces, but it will also bring new challenges—including whether AI itself is secure, resistant to prompt injection, and correctly understands on-chain operations. As interaction methods evolve rapidly, some issues will be resolved while new security challenges arise.