Experts polled by ForkLog are split over the present and future of decentralised autonomous organisations.
A few years ago, decentralised autonomous organisations (DAOs) seemed the first realisation of crypto-anarchist ideals: a future without hierarchies and bureaucracy, in which communities take decisions via smart contracts; code replaces the need to trust people; tokens apportion influence; votes displace politics. Something, however, went awry.
Cases of DAOs being paused or frozen are multiplying. Votes barely reach quorum, forums empty out, treasuries sit idle and projects wind down. No one declares a death; instead come euphemisms: “temporarily pausing governance”, “revisiting the model”, “optimising processes for speed”.
What happened? Has the idea crashed into the laws of human psychology, power and economics—or are we witnessing a crisis that will be followed by a powerful evolutionary refit of the system? We asked the experts.
Grasp a DAO—and die
Today more than 12,000 DAOs manage roughly $28bn in assets. Average turnout among decision-makers hovers around 20%, and in many cases only one in ten actually uses their right to vote. What initially promised radical decentralisation (one token—one vote; decisions by the community rather than a board-like cabal) looks very different in 2026.
In January, after its leader exited Scroll DAO, the organisation fully suspended operations amid uncertainty over which proposals were even under consideration—a curious look for a system that prizes transparency. Earlier, Jupiter froze all governance votes and locked access to its funds first until 2026, then 2027. Yuga Labs abandoned a DAO structure, citing inefficiency.
More often the story is less dramatic: a reversion to centralisation, where real power flows back to the developer team and the DAO remains a formality. Thus Compound drifted from an open system into a club of delegates.
Uniswap tries to preserve decentralisation (token voting and on-chain execution) but in practice leans towards centralising decisions via thresholds, filters and a delegation model—meant to ease governance overload and raise efficiency, but also narrowing the circle of true decision-makers.
A Cornell University group in 2025 published an analysis of Compound and Uniswap governance.
“Analysing more than 370 governance proposals and millions of on-chain events from inception to August 2024, we found substantial centralisation of voting power: as few as three to five voters were sufficient to sway most proposals. We also found that the cost of voting falls disproportionately on small tokenholders, and that strategic voting behaviour—such as delayed participation and coalition formation—further distorts governance outcomes. Our findings show that, despite their decentralised ideals, existing DAO governance mechanisms underperform in practice,” the authors reported.
Why so? The idea was elegant: give people governance tools—wallets, interfaces like Snapshot or Tally—and they will rush in. Every tokenholder could influence a project’s fate directly. But participation takes work: hours spent reading forum proposals, untangling tokenomics where yield farming, impermanent loss and gas fees interlock like a puzzle. Then comes voting on knotty trade-offs: “Should we raise token emissions to stimulate liquidity, or not?”
In reality, 1% of tokenholders control 90% of voting rights. This is not a bug so much as human nature: early zeal gives way to apathy and delegation—be it to an oligarch or to an AI.
DAOs have hit the limits of their architecture, of motivation, and of tokenomics. Roughly put, embed capital and you inherit the costs of capitalism. Our experts split into two camps: some say the idea has failed; others argue the model is simply evolving into more resilient forms.
DAO believers
DAOs have a habit of looking “effective” merely because they run many votes or boast high TVL. That does not mean they are well governed. Beyond voting, execution, transparency and the ability to allocate resources without constant hand-holding matter.
There are no universal yardsticks: DAOs vary widely by type—from grant-giving to protocol DAOs, from small clubs to billion-dollar treasuries. Still, some broad parameters apply.
Most assessments focus on four blocks:
- participation: how many people actually vote, how many delegates are active, what the quorum is and how votes concentrate;
- speed and execution: how quickly proposals move from discussion to execution, and how few decisions stall;
- economic outcome: whether the treasury grows, income exists and value accrues to participants;
- quality of governance: whether there is capture, burnout, factionalism or constant policy whiplash.
In short, judge a DAO by how swiftly and cleanly it turns collective choice into measurable outcomes while preserving transparency, durable decentralised participation and economic benefit for the protocol.
Has the DAO form proved itself inefficient by 2026? Web3 researcher Vladimir Menaskop answered unequivocally—provided we mean by “efficiency” the very things DAOs are made of: decentralisation, autonomy and organisation.
“As of 2026 DAOs are inefficient where they were created as an imitation of process, as an attempt to shirk responsibility (the peak here is the bZx case), rather than as an organically derived consequence of decentralisation as a principle. Take the largest lending protocol, AAVE: there the DAO, on the one hand, shields tokenholders from inefficient projects, and on the other, protects against hacks, because much is decided precisely at the DAO level. And so AAVE is effective. However, V4 moved towards centralisation, and the community immediately raised questions. A compromise has so far been found in the financial aspect—by sharing revenue—but that is only a temporary solution,” Menaskop claims.
In his view, Uniswap shows something similar: contentious debates erupt—over implementing BNB Chain, fees and payouts to tokenholders—but the system still moves forward. He classes the Bitcoin and Ethereum networks as effective too—“not classic firms but global teal corporations, for which a DAO is not just a form but the substance.”
Menaskop adds that the most effective DAOs are those that give communities more than a ballot:
“For example, AAVE executed a remarkable tokenomics overhaul towards not just staking but simultaneous buybacks; UNI moved from zero fees to positive ones; and SAFE, say, did the right claim for early users.”
DAOs have not proved their inefficiency; they have proved their limits. So argues Denis Smirnov of DAO Builders.
“I would describe what’s happening not as a DAO collapse, but as the end of their first, overly utopian version,” he says.
In his view, the first wave erred by reducing complex political coordination to a token-weighted vote.
“DAOs work where there is a shared resource, formaliseable rules and transparent execution: treasury management, grants, protocol parameter governance and the like,” Smirnov notes.
In February, Web3 and AI researcher, content strategist and Web3FuturePro founder Abubakar Yusuf Radda urged observers not to give up on DAOs:
“In 2026, decentralised autonomous organisations will not replace governments or corporations—they will offer a superior alternative for solving specific coordination problems: transparent funding, protocol governance, community participation. As legal recognition grows and tools improve, expect DAOs to appear in impact finance, open-source funding and even municipal pilot projects. The revolution is not top-down; it is distributed, on-chain and only just beginning.”
We are entering the second generation of DAO governance, Radda believes. The next wave will not abolish leadership, but it will force us to rethink it—and that matters. He highlights several trends to watch:
- hybrid governance (a DAO plus a legal entity);
- AI-assisted proposal evaluation;
- on-chain identity verification;
- governance-as-a-service platforms;
- decentralised autonomous organisations based on real-world assets.
He does not consider DAOs a panacea that erases politics, conflict and injustice. They enable programmable trust—a rarity in itself.
“Real change is not technological. It is philosophical. From centralised power to shared ownership. And this is possibly the most ambitious governance experiment of our time,” Radda argues.
The DAO undertakers
Others are convinced the experiment has failed.
Cyber~Congress co-founder Dmitry Starodubtsev says flatly that DAOs do not work. Avoiding power concentration, in his view, is not necessary at all.
“It’s more that there aren’t enough good leaders,” he concludes.
Allbridge.io co-founder Andriy Velykyy likewise speaks of a broken idea.
“The point is, DAOs emerged as a way to hand governance to the whole community. It turned out a crowd of macaques won’t write ‘War and Peace’. And they can’t and won’t govern anything. You either motivate that community very strongly, or that DAO will still dangle in the same way as a centralised system,” the expert says.
He pointed to a recent conflict between the AAVE community and developers over an attempt by the latter to seize control of the brand and revenues. Such scandals hurt projects and can kill them.
“In my own project, although I’m broadly sympathetic to decentralisation, I would be unlikely to hand governance of anything to the community. I believe in motivated voting. For example, if a protocol is already prepared to share some rewards with the community, to introduce some revenue sharing, profits, one can adopt a DAO scheme by showing that the most active will get more simply for clicking buttons and signing contracts,” Velykyy explains.
He also notes another motivation: in a number of projects DAOs were used primarily as a legal shield, not a real decentralisation mechanism. In recent years there has been a shift towards more traditional forms—plain incorporation, especially in jurisdictions where it is easier to build clear legal and operational models. That looks logical amid tighter regulation and the rising cost of complex blockchain structures.
Concentration of power: feature or bug?
That organisations created to eliminate centralised control often re-create it via tokenomics may be paradoxical, but is usually the natural outcome of development.
Denis Smirnov argues you cannot fully avoid concentration in token-based governance, which almost inevitably concentrates capital. But you can blunt it by splitting economic and governance rights, strengthening expertise-based delegation, adding liquidity pools and reputation systems, and moving away from “1 token = all power”.
Andriy Velykyy agrees. Avoiding concentration is nigh impossible, he says, because 90% of people simply won’t vote—out of apathy, caution or security concerns.
Even Vladimir Menaskop, who insists DAOs are “alive and kicking”, sees no way to avoid centralisation so long as voting is tied to token holdings. To him, that does not contradict a commitment to decentralisation.
“Voting is about democracy—that is, about the past; whereas a DAO is about anarchy—that is, about the present and the future,” the crypto-enthusiast says.
Menaskop cites a structural effect described by sociologists Robert Michels and Gaetano Mosca around the turn of the 20th century: any large organisation tends towards the rule of the few, and an organised minority almost always outmuscles a dispersed majority. That does not preclude successful DAOs. It means large communities should be split into smaller working groups, where decisions are faster and meatier.
That, he argues, is how DAOs really function: some questions interest only a narrow circle, while others—above all financial ones—require a wider audience. Small teams are thus critical, and resilience depends on whether grassroots decision-making remains real rather than ritual.
Often, many topics attract only a limited cohort of listeners, viewers and participants, while others—typically financial—draw the overwhelming majority. Hence DAOs need small groups of five to ten people.
“For Web3 projects this is native, and so 1inch pays tokenholders from a set of transactions, while UMA grants voting rights literally via email notifications. The grassroots level must be substantive, not for show. Then concentration of power, if not impossible, is at least unlikely,” Menaskop clarifies.
In short, concentration in token-based governance is a near-inevitable design effect, often harmful to transparency. When votes are yoked to tokens, power follows capital: those who amass tokens take most decisions; everyone else abstains or delegates to the same hyper-active few. The question today is less whether a DAO can eliminate centralisation, and more whether it can soften it through delegation, reputation systems, and separating economic from governance rights.
Can and should DAOs be “resuscitated”?
Revenue-sharing alone will not magically revive a DAO, but it can restore a token’s economic purpose—and give a frozen community a reason to re-engage. Andriy Velykyy accepts that logic, but adds a caveat:
“Tokens are basically dead right now—so what DAO are we talking about if 99% of projects on the market are stagnating with tokens and want to get rid of them? And a token with an extra function that you must go and apply proactively, rather than it working by itself—that, I think, is an illusion altogether.”
Menaskop is more optimistic about reviving frozen DAO structures:
“I myself am in a DAO that has already gone through three full cohorts: the first operated in 2016–2017, then a 2018–2019 team arrived, then there was a break—when the organisation functioned, and powerfully, but with a few leaders—and a new period began around 2023 and continues now. How it ends—we’ll see. But the fact is you can always breathe new life into both small and large DAOs.”
DAOs with vanishing participation can be revived only if a real object of governance remains, says Denis Smirnov—be it a treasury, a product, a protocol or a network of participants.
“I see the point of the new cycle in the pairing of DAOs with AI agents: they will remove routine, prepare proposals, run simulations and post-mortems, and take over part of operations. So I expect a DAO revival within the next year,” Smirnov forecasts.
Menaskop points to DAOs that have already been reborn, and highlights cases he considers most successful:
- BitDAO — the move to Mantle (L2) gave fresh impetus and evolution;
- ENS — spent a long time searching for its path and found it in a tacit pact with Vitalik Buterin (to stay off L2);
- SKY — its rebrand and relaunch can be deemed successful given the economic growth;
- Balancer — kept operating in a decentralised fashion after a hack.
To revive DAOs, he says, look to economics, ideas and community. New monetisation models can rekindle interest; without a compelling idea and ongoing product development, stagnation follows; without a strong community, a DAO loses resilience.
“DAOs are going through hard times: network support is being cut because infrastructure is costly or very costly; a number of projects are trying to revert to classic structures. Yet at the same time the DAO market is experiencing a period when innovation helps businesses grow in very difficult conditions, making it obvious to many that classic, old ways of governance must simply be discarded. And I am in favour of these changes,” the expert concludes.
A promising response to the governance crunch is a fundamental rethink of tokenomics—and, at a deeper level, of governance philosophy. Survivors will treat governance not as a formality but as a product, deserving as much attention as the technology—if not more.
Most experts today think DAO resilience will come from hybrid governance structures, reputation- and expertise-weighted voting, AI assistance in sifting data, and diversified treasuries. In any case, the experiment continues. Whether to join it or watch from the sidelines is up to each of us.