Vitalik Buterin: traditional on-chain apps may fade as AI agents take over
The Ethereum co-founder says bots could handle much of the network's activity within two years, with big implications for wallets, apps and security
Vitalik Buterin thinks the way people use Ethereum is about to change, and the change may not involve clicking buttons. The Ethereum co-founder suggested that traditional on-chain applications may decline as AI agents and bots increasingly handle transactions on the network within two years.
Buterin laid out his thinking at the OKX Now 2026 Summit in Singapore on October 6, 2026. His core argument: AI could become the new user interface for on-chain activity. Instead of navigating an app, a user would tell an AI agent what they want. The agent would then carry out the task directly on the blockchain.
He offered a personal example. Buterin said he updated his Ethereum Name Service (ENS) records in about five minutes by instructing a local AI agent, with no traditional user interface involved. ENS is the system that maps readable names to Ethereum addresses. Updating those records usually means opening a dedicated app, connecting a wallet and signing transactions. Buterin skipped all of that and just asked a machine to do it.
Buterin did not explicitly declare that on-chain apps will vanish. His remarks instead pointed toward a future where AI mediates much of what users do on the blockchain, with the app layer receding into the background.
The bigger Ethereum roadmap
The network is moving toward what he calls a “cryptographic world computer” model, which leans on verifiable proofs instead of repeating execution across the network: one party does the work, and everyone else checks a compact proof that it was done correctly.
Buterin has been talking about Ethereum’s overlap with AI since early 2026. His vision includes AI agents that take part in economic interactions, help verify transactions and strengthen security protocols. He has also discussed trustless and private AI interactions.
AI, tech, and the markets they move—in one daily briefing.
Daily. Free. Join 34,000+ readers across crypto, finance, and policy.
Looking further out, Buterin pointed to the Hegotá upgrade, planned for 2027. He described it as potentially Ethereum’s last fork that developers from the network’s 2015 era would still recognize.
The security problem nobody can prompt away
Buterin was not purely optimistic. He flagged several risks that come with handing transactions to software agents.
The first is privacy. Traditional pseudonymity on Ethereum could erode as AI systems handle more user activity. The second is prompt injection, when malicious instructions get slipped into the data an AI reads, tricking it into doing something its owner never asked for. An agent with signing power over a wallet makes that attack far more expensive to get wrong. The third is general AI agent security: if a bot can move funds, the bot itself becomes the thing attackers want to compromise.
Buterin also referenced a major incident that caused about $1.4 billion in losses by targeting Ethereum’s user-interaction layer. The place where humans meet the blockchain has already proven to be a weak spot. Swapping humans for agents does not automatically make that layer safer.
Buterin’s two-year horizon for AI-driven activity lines up loosely with the planned Hegotá upgrade in 2027. Buterin’s language was framed around what may happen, not what will. But when the person who co-founded the network updates his own ENS records by chatting with a local AI agent, it is a fair signal of where at least one influential voice thinks Ethereum’s user experience is heading.