The DAO Governance Crisis: Why You Cannot Code Away Human Nature

Christopher Louis Tsu, Chief Executive Officer at Venom Foundation

There is a particular kind of idealism that surfaces every decade or so in technology – the belief that a new tool will finally override the messier aspects of human behaviour. In the 1990s it was the open internet that would flatten hierarchies and democratise knowledge. In the 2010s it was social media that would empower civil society and hold power to account. And in this decade it has been the DAO, the decentralized autonomous organization, that was supposed to deliver transparent, fair, and incorruptible governance to digital communities. On paper, it sounds perfect. In practice, it runs headfirst into a wall that no smart contract can breach: human nature.

I have spent four decades building companies across electronic design, satellite communications, medical biotech, and now blockchain infrastructure. That is long enough to have seen many revolutionary ideas arrive with great fanfare and quietly settle into something far more modest than their evangelists promised. DAOs are following the same trajectory, and I think it is time we talked honestly about why.

The numbers tell a stark story. According to Snapshot Labs data, the average voter turnout across DAOs remains below twenty percent. Even significant protocols like Maker and Uniswap struggle to attract more than ten percent participation on critical proposals. A peer-reviewed study published in the Journal of Finance and Data Science examining governance on Compound, Uniswap, and Ethereum Name Service found that the majority of voting power is concentrated in a small number of addresses. Research from Cornell and the National University of Singapore, presented across multiple academic venues in 2025, found that the top decile of voters controls 76.2 percent of voting power in a typical proposal – a concentration that surpasses what we see in traditional corporate governance. Chainalysis has reported that in ten major DAO projects, just one percent of all holders controlled ninety percent of the vote.

So here we are. A system designed to eliminate plutocracy has, in many cases, reproduced it with algorithmic efficiency.

The DAO faithful will tell you this is simply an early-stage problem, an awkward adolescence that better mechanism design will fix. Liquid democracy, quadratic voting, reputation-based systems, vote-escrowed tokens – the solutions are always just one protocol upgrade away. And some of them are genuinely clever. But they all share a fundamental blind spot: they assume that participants will play by the spirit of the rules, not merely the letter. Four decades of watching markets, boardrooms, and geopolitics have taught me that this assumption is, at best, naïve.

Let me offer an analogy that I think clarifies the problem rather well. I am from Switzerland. The grand offices of the United Nations used to be just down the road from where I grew up. Over the decades I have met and socialised with a fair number of people from that world. The UN established committees, rules, conventions – a beautiful architecture of international law designed to ensure cooperation and prevent conflict. On paper, it is the ideal world order. Everyone should follow those rules. Until one African state invades another and shoots whoever is in the way – enemy, civilian, or blue-helmeted UN soldier. Bang. Get out of my way, I have the guns. When Russia invaded Ukraine, the UN held meetings and issued press releases that read like statements from privileged university students, spent millions broadcasting perfect ideals and preaching international rule of law. But absolute power is absolute. Similarly, NATO stood behind a microphone with no meaningful army for years. Along came Trump and punched them in the face. Power is ultimate power, and humans always want control.

DAOs are ideal on paper. They are fair, transparent, and perfect decision-making systems – in much the same way that the UN Charter is a perfect document, or the World Trade Organization’s rules are perfectly logical. Until someone cheats. China games the WTO because it can. The United States sanctions Russia because it can. A whale in a DAO will manipulate the vote because he can. This is raw power, and no governance token redesign will change the underlying incentive.

A group of humans will club together and aggregate their votes to game the system, because coordination for self-interest is one of the oldest behaviours in our species. An institutional investor will not entrust capital to a DAO – not because it is not the fairest structure on offer, but because they want to know their person on the inside has their best interests on his list. You cannot go against hardwired human behaviour. We have been running this operating system for a very long time, and no software update from a crypto protocol is going to override it.

The evidence is not merely theoretical. In April 2022, an attacker used flash loans to borrow over one billion dollars’ worth of tokens and seized 182 million dollars from Beanstalk’s treasury in a single Ethereum block. Everything was technically decentralized. Everything was technically governed by the community. And everything was technically stolen. The CFTC obtained a default judgment against Ooki DAO in 2023, with the court agreeing that the DAO was a “person” and an unincorporated association – a ruling that effectively held every token holder liable. As Bloomberg’s Matt Levine put it at the time, it is possible that DAOs are the worst of all worlds: their tokens are similar enough to corporate shares to be subject to securities laws, but different enough to create unlimited liability for their holders.

The institutional response has been telling. Jupiter, the largest decentralized exchange aggregator on Solana with over two billion dollars in deposits, paused all DAO governance voting in mid-2025 after its leadership acknowledged that the structure was not working as intended. Yuga Labs scrapped its ApeCoin DAO structure entirely. Major protocols are quietly reverting to foundation-led decision-making or delegating authority to small committees. In the Arbitrum ecosystem, the CEO of Offchain Labs addressed community complaints about decentralization by noting that the DAO had confused decentralization with direct democracy – and that direct votes on every operational matter were actively harmful.

I recognise this pattern because I lived through a nearly identical one with the broader decentralization thesis. I remember crypto people going on and on about how decentralization would take over the world and make it a better place. Banks were going under. Crypto would kill banks. Decentralization would lift the middleman up. I was always the grumpy old man saying, “Shut up, kid, that will not happen until regulation is in place”, and by that time this technology will be an institutional-grade product – and we will be back to the same show.That is precisely what happened. Bitcoin ETFs are managed by BlackRock. Crypto custody is offered by Fidelity. The revolution got absorbed by the establishment, as revolutions in finance always do.

DAOs will follow the same arc. Decentralization had, and still has, genuine limited use cases. It helps balance things out. But it never did and never will take over the world. The same applies to DAO governance. There will be contexts where it adds real value – community treasuries for open-source software, perhaps, or coordinating small and highly aligned groups around specific missions. Wyoming’s DUNA framework and the SEC’s evolving posture show that regulators are trying to accommodate the concept, and that is constructive. But the vision of DAOs as a universal governance model – transparent democracy at scale, replacing boards and executives and representative structures – is a fantasy that collapses the moment it meets the reality of concentrated capital, rational apathy, and the human appetite for control.

The organisations that will actually succeed in Web3 will converge on hybrid models. They will use on-chain transparency for treasury management and specific, well-scoped decisions. They will delegate operational authority to accountable leadership – real people with real names who can be held responsible. They will maintain the ethos of community input without the paralysis of community consent on every operational detail. In short, they will look remarkably like well-governed traditional organisations that happen to use blockchain tooling where it provides a genuine advantage.

DAOs are not the nirvana for governance any more than the United Nations is the arbiter of world peace. Both are useful constructs that serve real purposes within strict boundaries – and both collapse the moment someone with sufficient power decides to ignore the rules. The industry should build governance structures that work for the world as it is, not the world as we wish it were.