Ethereum co-founder Vitalik Buterin is calling for a fundamental redesign of decentralized autonomous organizations, warning that most DAOs today are failing to deliver on their promise of better governance than traditional companies or political systems.
In a recent post on X, Buterin argued that many DAOs have become little more than onchain treasuries controlled by token-weighted voting, a model that he believes is inefficient, easy for large holders to dominate, and poorly suited to solving real governance problems.
Why Vitalik Buterin Thinks Today’s DAOs Are Broken
According to Buterin, the dominant DAO structure copies the same pattern across the ecosystem. Token holders vote on proposals, funds are distributed from a shared treasury, and governance is treated as a box-ticking exercise rather than a core product.
He warned that this approach leaves DAOs vulnerable to “capture” by whales and fails to address the weaknesses of human decision-making that decentralization was supposed to improve. Low voter turnout, decision fatigue and social pressure have become persistent problems, even as DAO market capitalization has grown to at least $17.5 billion, based on CoinMarketCap data.
Buterin believes the next generation of DAOs must move beyond generic governance and focus on solving specific infrastructure challenges that the crypto ecosystem depends on.
A New Blueprint: From Oracles to Onchain Dispute Resolution
Instead of treating governance as an afterthought, Buterin says DAOs should be purpose-built for concrete problems.
He highlighted three priority areas:
- Better oracles, which determine how off-chain data is brought onchain
- Onchain dispute resolution, including subjective cases such as insurance claims
- Long-term project stewardship, ensuring critical infrastructure survives even if founding teams disappear
He also proposed shared registries such as anti-scam lists and standardized frameworks that make it easier to spin up short-lived funding vehicles for specific missions.
In his view, DAO design and communication layers should account for roughly half of a project’s effort, not a small add-on after the product is built.
Concave vs. Convex Governance: Different Problems Need Different DAOs
Vitalik Buterin framed his thinking using a “concave vs. convex” governance model.
For concave problems, where compromise leads to better outcomes, DAOs should maximize robustness by aggregating input from many participants. Examples include public goods funding or ecosystem standards, where diversity of opinion improves results.
For convex problems, where bold, decisive bets matter more than consensus, Buterin said stronger leadership is often necessary. In these cases, decentralization should exist to hold leaders accountable rather than replace them entirely.
This hybrid approach, he argues, is more realistic than forcing every decision into a one-token-one-vote framework.
Privacy, AI, and the Fight Against Decision Fatigue
Buterin also warned that governance will fail unless DAOs fix privacy and participation issues.
Without privacy, voting becomes a social game where pressure and signaling distort decisions. He pointed to technologies such as zero-knowledge proofs, secure multi-party computation and fully homomorphic encryption as tools that could protect voter privacy while preserving transparency.
To reduce decision fatigue, he suggested that artificial intelligence could assist with analysis or allow users to delegate voting to locally controlled models, while cautioning against letting AI run DAOs outright.
As Ethereum’s ecosystem matures, Buterin’s message is clear. If DAOs want to outperform traditional institutions, governance must become a serious engineering discipline, not a copy-paste template.
Read Also: Hong Kong Crypto Industry Seeks Softer OECD CARF Tax Reporting Rules
