Ethereum co-founder Vitalik Buterin has called for a rethink of decentralized autonomous organization (DAO) design, arguing that most DAOs remain stuck in a shallow model that fails to fix the weaknesses of politics and corporate governance. In a Monday post on X, Buterin said today’s DAO blueprint often amounts to “a treasury controlled by token holder voting,” and the sector needs to move beyond that default.
Buterin warned that the standard treasury-and-token-vote structure has spread widely because it is simple to deploy. However, he said it delivers weak outcomes. It breeds inefficient decision-making, is liable to governance capture, and fails to deliver improved fairness or accountability. His perspective is that DAOs must migrate towards governance infrastructure rather than mere tokenized treasuries.
Buterin outlined a broader vision in which DAOs take responsibility for infrastructure that requires collective coordination and credible neutrality. He highlighted three major targets: better oracles, on-chain dispute resolution, and long-term project stewardship.
Oracles often decide what “truth” enters a blockchain system, which makes them central to DeFi, insurance, and settlement. On the other hand, the matter of dispute resolution is relevant whenever there is a need for subjective assessment in those agreements that happen on-chain, for instance, in insurance disputes, fraud cases, or in matters of the delivery of services. According to Buterin, DAOs can solve these challenges by proper design in their governance instead of relying solely on token votes for legitimacy.
He also pointed to governance use cases that demand continuous curation. DAOs could maintain shared lists such as anti-scam registries, build common formats for short-term funding vehicles, and preserve protocols after their original teams disappear. These functions matter for crypto’s long-term resilience, especially as many projects lose contributors after early hype cycles fade.
Buterin framed his argument around his earlier “convex vs concave” governance lens. He said different types of problems require different kinds of governance.
For concave problems, compromise is better than a coin flip. These decisions benefit from wide input, diverse views, and high robustness. In this case, DAOs should focus on aggregation mechanisms that pull signals from many participants while resisting manipulation.
For convex problems, a decisive bet can outperform slow consensus. Buterin said DAOs should allow strong leadership in these cases. Rather than attempting to supplant leadership through decentralization, DAOs can utilize the concept of decentralization to make leaders accountable.
This approach has implications because, in it, one-size-fits-all governance does not work. Buterin wants developers to consider the architecture of governance as something serious in the area of engineering.
Buterin also warned that DAOs will fail unless they solve two structural issues: privacy and participation fatigue.
Without privacy, governance turns into a social game. Participants vote based on status, influence, and pressure rather than honest preferences. Buterin pointed to zero-knowledge proofs as a core tool to protect voter privacy. He also mentioned secure multi-party computation and fully homomorphic encryption as possible approaches in some cases.
Decision fatigue creates another major failure mode. If DAOs require constant voting, turnout collapses after early enthusiasm. Buterin suggested that AI could help reduce fatigue by supporting analysis or enabling delegation to locally controlled models. However, he cautioned against allowing AI to run DAOs directly.
Buterin’s comments land as DAO ecosystems expand while still struggling with low turnout and token concentration. DAO tokens now hold a market capitalization of at least $17.5 billion, but many governance systems still rely on a narrow set of active voters.
Buterin argued that teams building new governance systems should treat DAO design and the communication layer around it as “50% of their job, not 10%,” especially if Ethereum’s decentralized base layer is meant to carry into the applications built above it.
Highlighted Crypto News:
Canaan Risks Nasdaq Delisting as Shares Slip Below $1 Again

