Ethereum's Co-founder Vitalik Buterin Suggests AI Stewards Could Transform DAO Governance

Ethereum co-founder Vitalik Buterin has proposed the integration of AI 'stewards' into DAO governance, aiming to address challenges like low participation and power centralization by enabling AI agents to vote based on users' values and past interactions. This innovative model, employing privacy technologies like multi-party computation and zero-knowledge proofs, aims to enhance secure and anonymous voting within decentralized frameworks.

Chris Wilson

February 21, 2026

Vitalik Buterin, Ethereum's co-founder, recently unveiled a novel proposal aimed at revamping how decentralized autonomous organizations (DAOs) operate. By integrating personal AI agents into DAO governance, Buterin suggests a path forward that mitigates issues of low participation and centralization of power that have plagued DAO structures.

Buterin’s concept revolves around AI 'stewards' that would autonomously vote on behalf of their human counterparts, relying on a deep learning model that understands individual user values and past interactions. These AI systems, which Buterin detailed on the social media platform X, would function within privacy-preserving frameworks such as multi-party computation (MPC) and trusted execution environments (TEEs). This setup ensures that sensitive data remains shielded while facilitating secure, private voting as reported by CoinDesk.

Additionally, Buterin addresses the importance of maintaining voter anonymity through zero-knowledge proofs (ZKPs), a cryptographic method that allows users to verify voting eligibility without disclosing their identities or the specifics of their vote. This layer of security is critical in preventing undue influence, coercion, or the strategic copying of votes from influential token holders.

The AI stewards aren't just automated clerks; they're designed to engage intelligently with the flood of proposals that DAOs encounter, sifting through potential spam using prediction markets where bets on proposal success could regulate the quality of submissions.

This approach could dramatically scale DAO participation while ensuring that governance remains in the hands of the many rather than the few. However, the success of this model hinges on the robustness of AI training and the security of the underlying technological frameworks. As Buterin himself notes, these AI systems must be sufficiently advanced to interpret complex human values and nuanced governance issues-an ambitious goal that intertwines cutting-edge AI with emerging blockchain technologies.

While the proposal is still theoretical, its implementation could mark a significant evolution in how digital communities govern decentralized networks and resources. If proven effective, AI stewards could indeed herald a new era of democratic, efficient, and secure digital governance, transforming the foundational mechanics of DAOs.

Sign up to Radom to get started