While DAOs were meant to embody the promise of decentralized governance, Vitalik Buterin today delivers a harsh assessment : their current model is exhausted. In a widely shared post, the Ethereum co-founder denounces rigid structures, dominated by large holders and unable to address complex coordination challenges. His call for a new design marks a pivotal moment for DAOs, urging them to move beyond simple voting logic to become true on-chain infrastructures.
In a message published on January 19, Vitalik Buterin was particularly critical of the evolution of DAOs, denouncing a model that he considers “rigid, excessively tokenized, and vulnerable to the perverse effects of large holders”, while a controversial vote stirred the Aave protocol community.
In his view, the initial promise of decentralized coordination was hijacked by an almost exclusive focus on token-based voting, at the expense of building resilient institutions. “Many DAOs today are treasuries managed by snapshot and a few multisigs”, he writes, pointing to a loss of substance compared to original ambitions.
To illustrate this drift, Buterin highlights several systemic flaws observed in many emblematic DAOs :
This severe diagnosis is accompanied by a warning. Without structural transformation, DAOs risk losing their relevance and their ability to embody a genuine alternative to centralized models.
To escape this deadlock, Vitalik Buterin puts forward several proposals that sharply differ from the dominant model. He advocates for DAOs that move away from purely financial logic, to play a role as technical infrastructure within the ecosystem.
According to him, these organizations should focus on functions such as oracles, on-chain dispute resolution, or coordinating projects limited in time. “We need more DAOs, but different and better DAOs”, he summarizes on X.
Among the proposals, Buterin emphasizes the importance of privacy technologies to reform governance: encrypted voting, zero-knowledge proofs, and secure multiparty computation.
He also mentions the use of artificial intelligence, not to replace human decisions, but to reduce the complexity of internal processes. By introducing “lenses of governance” adapted to the decision-making context, he proposes modular governance capable of adapting to the nature of decisions (technical or political, urgent or deliberated).
These proposals are not only aimed at improving the technical efficiency of DAOs. They represent a conceptual shift towards structures capable of outliving their creators and taking on key roles in DeFi, insurance, or decentralized identity management. If they resonate with developers and founders, these ideas could inspire a new generation of DAOs better equipped to meet the political, economic, and social challenges looming in Web3.
