【Vitalik proposes that Ethereum will use AI to verify the security of entire programs】

Ethereum co-founder Vitalik has publicly outlined the latest technical roadmap for AI and cybersecurity. He rejected the view that AI will completely destroy cybersecurity, pointing out that security defenses have inherent advantages, with the core lying in proactively writing and verifying code that is more resistant to attacks.

Vitalik stated that in the future, AI should aim to verify entire programs, including databases, networks, and caches, rather than only checking critical parts. He made an analogy: if AI can prove the Navier–Stokes equations and Fermat’s Last Theorem, it should also be able to formalize “the program is secure” as a mathematical theorem—thereby covering attack scenarios such as forgery, replay, server intrusion, key leakage, and side-channel attacks.

He defined this direction as a development priority for Ethereum in the coming years. This statement is intended to strengthen Ethereum’s long-term technical narrative in the field of smart contract security, emphasizing improved protocol-level safety through formal verification. The market should pay attention to whether this direction translates into concrete on-chain verification tools or ecosystem applications—this will be a key observation point for assessing the execution strength of Ethereum’s technical roadmap.