Ethereum co-founder Vitalik Buterin said AI is showing capabilities that go well beyond basic assistance, including breaking out of sandboxes, attacking websites, finding software vulnerabilities, helping the Ethereum protocol identify bugs, implementing complex cryptography, and carrying out formal verification work. He also shared a recent personal example: about a month ago, he updated ENS records for the first time without using a traditional user interface, asking only a local AI agent to write the script. The task was completed in about five minutes. Based on that experience, Buterin said he expects AI-driven execution of complex onchain operations to gradually become normal and to serve as a new user interface in more scenarios. He added that the roughly $1.4 billion in losses tied to the Safe-related hack were aimed at the interaction layer between users and onchain systems rather than smart contracts themselves. In his view, having AI handle transactions directly could reduce risks introduced by web interfaces, though it would also bring a new set of issues, including AI security, prompt injection, and whether the system correctly understands the intended action.
Odaily reported that Ethereum co-founder Vitalik Buterin said AI is showing the ability to break out of sandboxes, attack websites, discover software vulnerabilities, help the Ethereum protocol find bugs, implement complex cryptography, and perform formal verification.
Buterin said that about a month ago, he updated ENS information for the first time without using a user interface. Instead, he asked a local AI agent to write the necessary script, and the operation was completed in about five minutes.
AI as a new interface layer
He said he expects AI to directly carry out complex onchain operations on a regular basis over time, with AI becoming a new user interface across more use cases.
Interaction-layer risk and new security issues
Buterin also said the Safe-related hack caused about $1.4 billion in losses, and that the attack targeted the interaction layer between users and onchain systems rather than onchain contracts. He said direct AI handling of transactions could reduce risks introduced by web interfaces, but would also create new challenges around AI security, prompt injection, and the accuracy of operational understanding.
This article was originally published by Bit.Fan. For more cryptocurrency news and market insights, visit www.bit.fan. Disclaimer:
The market information, project data, and third-party content displayed on this platform are for industry information sharing only and do not constitute any form of investment advice or return commitment.
Cryptocurrency trading carries high risks. Users should fully assess their risk tolerance and make independent decisions. All profits, losses, and legal responsibilities are borne by the users themselves.
