Vitalik Buterin says artificial intelligence could help build more effective decentralized governance systems and empower users to make better-informed decisions.
In a post on X on Sunday, the Ethereum co-founder argued that one of the core challenges facing democratic and decentralized governance models, such as decentralized autonomous organizations (DAOs), is the “limits to human attention.”
He noted that many governance decisions require specialized knowledge or significant time commitments — resources most participants lack.
“The usual solution, delegation, is disempowering; it leads to a small group of delegates controlling decision-making while their supporters, after they hit the delegate button, have no influence at all,” Buterin wrote.
Participation in decentralized autonomous organizations (DAOs) is estimated to average between 15% and 25%, a dynamic that can contribute to power concentration and weaker decision-making. In extreme cases, low engagement can open the door to governance attacks, where a malicious actor accumulates enough tokens to push through harmful proposals without broader community awareness.
AI-powered assistants could vote on users’ behalf
Vitalik Buterin suggests that personal assistant large language models (LLMs) could help address this “attention problem” by equipping users with relevant context and analysis before votes.
“If a governance mechanism depends on you to make a large number of decisions, a personal agent can perform all the necessary votes for you, based on preferences that it infers from your personal writing, conversation history, direct statements,” he explained.
Such AI tools, he argues, could allow participants in ecosystems like Ethereum to remain meaningfully involved in governance without needing to devote extensive time or specialized expertise to every proposal.
“If the agent is unsure how you would vote on an issue, and believes the matter is important, then it should ask you directly and provide all relevant context,” Vitalik Buterin added.
Lane Rettig, a researcher at the Near Foundation focused on AI and governance, previously told Cointelegraph that the nonprofit was exploring a similar concept — AI-powered “digital twins” capable of voting on behalf of DAO members to address low participation rates.
Privacy remains a key concern
Buterin also noted that highly decentralized governance models face challenges when important decisions rely on private or sensitive information, such as negotiations, internal conflicts or funding allocations.
“Typically, orgs solve this by appointing individuals who have great power to take on those tasks,” he said.
As an alternative, he suggested a system in which users submit their “personal LLM into a black box.” In this model, the AI would access confidential information, make a determination and output only its final judgment — without revealing the underlying private data to the user or others.
“All of these approaches involve each participant making use of much more information about themselves, and potentially submitting much larger-sized inputs. Hence, it becomes all the more important to protect privacy,” Buterin said.
