h hoge.gg
Subscribe
BTC$67,432.18+2.34%ETH$3,521.44+1.08%SOL$178.62-0.62%BNB$612.30+0.41%XRP$0.6234-0.18%ADA$0.4521+3.12%DOGE$0.1623+1.86%AVAX$38.71-1.24%LINK$17.84+0.92%HOGE$0.00004120+4.21%
BTC$67,432.18+2.34%ETH$3,521.44+1.08%SOL$178.62-0.62%BNB$612.30+0.41%XRP$0.6234-0.18%ADA$0.4521+3.12%DOGE$0.1623+1.86%AVAX$38.71-1.24%LINK$17.84+0.92%HOGE$0.00004120+4.21%
● Security & Exploits

The Great Retreat: When the DAO Vote Became a Liability

In 2026 the token vote turned into both an exploit and a legal liability, so crypto's biggest DAOs handed power back to foundations, wrappers and founders. Inside the great governance retreat.

The $1.5 Billion That Walked Out of Sky

On the morning of 19 November 2026, a single holder linked to the Poloniex exchange pulled nearly two-thirds of the available reserves out of Sky’s Peg Stability Module. Within a day, more than $1.5 billion in USDC had left the module, and the market value of Sky’s USDS stablecoin had fallen by over 30 percent to about $3.2 billion, even as the token held its $1 peg, CoinDesk reported. A bank run, in other words, triggered not by insolvency but by a compliance question.

The trigger was a report from crypto risk firm Inca Digital, which traced $518 million in transactions through the module to 422 wallets tied to hackers, sanctioned entities and other illicit actors, including $101.9 million linked to North Korean state hackers. Some of it came from the $1.5 billion Bybit hack of February 2025: within ten days of that theft, $91.5 million moved through the module, and nearly $30 million of it after the FBI publicly blamed North Korea, according to CoinDesk’s account of the investigation.

The module itself is dull plumbing. It lets anyone swap Circle’s USDC for Sky’s USDS or DAI, which moves dollars out of a freezable stablecoin and into one that cannot be frozen. The danger is where those dollars sit: roughly $2.7 billion of USDC in a single Ethereum address held through Coinbase Prime. If a court ordered Circle to blacklist that address, the whole reserve would lock at once, freezing the legitimate money backing more than $8 billion of DAI and USDS right alongside the illicit deposits. “I think this is a low probability outcome, but it is not entirely hypothetical,” Amey Dandawate, a ratings director at Bluechip Ratings, told CoinDesk.

Sky founder Rune Christensen pushed back, noting that USDS “remains over collateralized with ~$10B in backing” and that the protocol was operating normally, and the firm said the flagged wallets amounted to roughly 0.03 percent of the module’s $968 billion in cumulative volume. But the detail that matters for this story is who could have stopped the illicit flows, and who a regulator would hold responsible for them. Sky is not a licensed money transmitter with a compliance officer. It is governed by MKR and SKY token holders, who vote on exactly these parameters and can modify or halt the module through a governance vote. “Calling your protocol ‘decentralized’ while retaining admin keys, collecting fees, and pushing upgrades does not make it decentralized in the eyes of sanctions law,” legal analyst Ari B. Good told CoinDesk.

The episode is a near-perfect specimen of the year’s central problem. Sky is one of the largest and most battle-tested protocols in DeFi, run by people who are neither reckless nor anonymous, and it was exposed precisely because its governance is real: token holders genuinely can move the parameters a regulator cares about. Decentralization did not dilute the responsibility. It spread it across a crowd of voters, many of whom may not have realized they had signed up to be controllers.

The Vote Became Two Things at Once

This publication has spent 2026 documenting one half of that sentence: the governance vote as an exploit. In case after case, an attacker bought a thin, cheap token, passed a malicious proposal and walked the treasury out the front door, with the smart contracts executing exactly as designed. BonkDAO lost about $20 million that way in July, when seven wallets controlled almost the entire vote, per crypto.news. Term Finance lost 68 percent of its vaults in August to an attacker who, as PeckShield and CertiK put it, “targeted the voting mechanics rather than any flaw in the underlying smart contract code,” Crypto Briefing noted. On 24 October, an attacker took an Aave flash loan, seized two of three “Senate” seats at a protocol called X100 and drained the treasury in a single Ethereum transaction, a KuCoin report put the loss at $4.85 million.

The Sky scare is the other half: the vote as a liability. If a handful of addresses can reconfigure a system that is moving sanctioned money, then those addresses are the controllers, and control is what attaches legal responsibility. The two halves point the same way. When your governance token is both the cheapest route for a raider to seize your treasury and the clearest path for a prosecutor to find a defendant, the rational move is to make the token control less. That is what the biggest names in crypto spent 2026 doing.

What makes 2026 the hinge is that both pressures arrived at full force in the same year. The summer takeover wave proved the attack was not a one-off but a repeatable business, and the autumn dragged the liability question out of academic papers and into a live stablecoin run. A protocol can shrug off a single bad headline. It is much harder to keep an open, capturable vote when that vote is at once the cheapest way to lose your treasury and the surest way to hand a prosecutor a defendant.

Over roughly eighteen months, Balancer voted itself out of existence, ENS handed its treasury to a foundation, Aave consolidated power around its founding company, ApeCoin dissolved its DAO into a Cayman entity, and Uniswap wrapped its governance in a Wyoming legal shell. The leaderless-organization dream, the thing that was supposed to make crypto different, recoiled into foundations, wrappers, councils and bosses. The uncomfortable part, for anyone who believed the dream, is that the retreat is rational.

Why the Vote Is So Cheap to Steal

To see why protocols are fleeing their own governance, start with the math of attacking it. Andreessen Horowitz researchers laid it out in a now-standard framework: an attack pays whenever the value of the attack exceeds the cost of acquiring the votes plus the cost of executing it, as their essay on DAO governance attacks explains. When a governance token is thinly traded, or its price has collapsed, the cost of acquiring votes falls toward zero while the treasury those votes control does not, and the inequality flips in the attacker’s favor.

The deck is stacked before anyone shows up. A 2022 Chainalysis study of ten major DAOs found that fewer than 1 percent of holders controlled about 90 percent of the voting power, Cointelegraph reported. A 2026 analysis of 48 large Ethereum DAOs, cited by CryptoSlate, found the top ten holders controlling more than half the votes in 39 of them, with Convex commanding roughly 53 percent of Curve’s votes and Aura about 65 percent of Balancer’s. Turnout, meanwhile, is usually in the single digits. Concentration plus apathy means a motivated buyer rarely needs much.

Ethereum co-founder Vitalik Buterin diagnosed the root cause years ago: a governance token bundles two rights, an economic interest in the protocol and a vote over it, and those rights are, in his words, “very easy to unbundle,” as he argued in an essay on moving beyond coin voting. A flash-loan borrower, or a holder who is shorting the token, can wield governance power with no lasting economic stake, which is precisely the person you least want deciding where the treasury goes. His verdict on the usual fix is just as bleak: timelocks, he has written, are “more like a paywall on a newspaper website than they are like a lock and key.” The economics are familiar to anyone who has watched value extracted on-chain through MEV; a governance vote is simply another lever for moving money that was never yours.

The cost side of that equation keeps falling for reasons that have nothing to do with any single protocol. A token unlock that floods the market, or a bear phase that halves a governance token’s price, cuts the dollar cost of buying a majority of the vote while the treasury those votes control stays exactly as large. 2026 was thick with unlock schedules, and each one nudged some thinly held governance token closer to the point where seizing it costs less than the prize inside.

Balancer: A DAO Votes Itself Out of Existence

In September, holders of BAL, the token behind one of DeFi’s oldest automated market makers, did something DAOs almost never do on purpose: they voted to shut the whole thing down. Proposal BIP-928, an “orderly winddown,” passed with more than 99 percent of the roughly 17.2 million BAL cast, clearing the 5 million quorum when the Snapshot vote closed on 29 September, according to the Balancer governance forum.

The cause was not a single catastrophe but a slow bleed. Protocol revenue had fallen from about $97,000 in June to $30,000 in August against a monthly burn near $150,000, and a $128 million exploit of a Balancer v2 pool in November 2025 had damaged the brand even though the newer v3 code was architecturally separate. “With a working product suite, a lean structure and full effort from everyone who stayed, the traction the plan required did not come,” wrote Marcus Hardt, a former Balancer Labs chief executive who put the proposal forward.

The plan is a liquidation, not a pivot. Pools moved to withdrawals only on 30 October, the v3 vault pauses at the end of November, and the roughly $9 million treasury goes back to BAL holders who burn their tokens, distributed in kind over a schedule that opens in May 2027 and ends with a final sweep in July 2028, The Crypto Times reported. Yahoo Finance put the eventual payout at about 16 cents per BAL. A DAO’s most decisive governance act, it turned out, was to stop governing.

ENS: From Protocol DAO to Foundation

The Ethereum Name Service took a gentler road to a similar place. In August, ENS tokenholders voted to hand administrative control of the project’s roughly $65 million endowment to the newly formed ENS Foundation and its small board, a change approved with about 70 percent support, The Block reported. The DAO kept the right to appoint and remove directors, but the day-to-day stewardship of the money moved to a professional body.

The justification was almost clinical. “A DAO, on its own, does not have the legal and operational capacity to act in many of the venues where that infrastructure now needs representation,” the proposal argued. A token vote cannot sign a contract, open a bank account, keep a lawyer on retainer or stand in a courtroom. A foundation can. The pattern, a decentralized project discovering it needs a center after all, is not unique to ENS; Bittensor promised no center and a corporate parent built one anyway.

ENS also spent the summer in a civil war over who controls its security council, with co-founder Nick Johnson using roughly 80 percent of the votes cast to block one renewal before a new eight-member council was finally approved. The fight was vicious precisely because everyone understood the stakes: whoever holds the emergency brake holds the protocol. Moving real authority to a foundation is the same move made quietly, through a reshuffle rather than a brawl.

Aave: Re-Centralization by Attrition

Aave reached the same destination without a single dramatic vote. Over the first half of 2026, the lending protocol shed its entire layer of independent service providers: BGD Labs left in February, the Aave Chan Initiative announced its wind-down in March, and risk manager Chaos Labs exited around April, as CoinDesk chronicled.

What drove them out was a governance-power problem. “The temp check demonstrated that a single entity holds enough voting power to pass its own budget proposals over community opposition,” said Marc Zeller, founder of the Aave Chan Initiative, as The Block reported. By the time Aave v4 launched on Ethereum mainnet at the end of March, the protocol, with deposits around $26 billion, had effectively re-centralized around its founding company, Aave Labs, not by dissolving the DAO but by hollowing it out until the only coordinated bloc left was the one aligned with the founder.

Analysts reached for a blunter word: privatization. When the independent risk managers and contributors who once checked a founding team all head for the door at once, what remains is a DAO in name whose real decisions are set by one company and merely ratified on-chain. For a depositor the practical question becomes whether that company is competent and aligned, which is the same question you would ask of any bank, and nothing like the trust-minimized promise the token was sold on.

ApeCoin and Jupiter: The Founder Comes Back

Not every retreat was dressed up as a restructuring. In June 2025, ApeCoin holders approved AIP-596, “Sunsetting the DAO and Launching ApeCo,” transferring roughly $168 million in assets to a new Yuga-controlled Cayman entity with close to 99 percent support. Yuga Labs co-founder Greg Solano had made the case bluntly, calling the DAO’s work “sluggish, noisy, and often unserious governance theater” that funneled money into “vanity proposals and low-impact initiatives,” CoinDesk reported. The community voted, almost unanimously, to be governed by a company instead.

Solana’s largest exchange aggregator, Jupiter, reached for the pause button rather than the exit. It suspended DAO voting in June 2025, citing a “breakdown in trust” and plain community fatigue, CoinDesk noted. Voting resumed in early 2026, but the episode made the point: once the townhall becomes a liability, founders feel entitled to switch it off.

There is a dark irony in the mechanism. In almost every case the retreat was itself ratified by a vote: ApeCoin holders voted to abolish their DAO, Balancer holders voted to wind theirs down, Uniswap holders voted to wrap theirs in a Wyoming shell. The last meaningful act of many token-governed communities in 2026 was to use the vote to vote the vote away, a collective decision that collective decisions were not working.

Put together, the cases rhyme. Each protocol concluded that an open token vote was more trouble than it was worth, and moved power somewhere more accountable, or at least more legible to a regulator and a court.

ProtocolWhat it didMechanismWhere power went
BalancerDissolved the DAOBIP-928 winddown; treasury burned back to holdersNowhere; the protocol is closing
ENSMoved the endowment to a foundationTokenholder vote, August 2026ENS Foundation board
AaveRe-centralized by attritionService providers quit; a founder-aligned bloc dominatesAave Labs
ApeCoinDissolved into a companyAIP-596, about 99 percent approvalApeCo, a Cayman entity
UniswapWrapped governance in a legal shellAdopted a Wyoming DUNA, named DUNILegal entity plus foundation
JupiterPaused the voteGovernance suspended in 2025, later resumedCore team

Governance Minimization: The Case for Less Democracy

Behind the individual retreats sits an intellectual argument that has quietly become orthodoxy. a16z Crypto general counsel Miles Jennings calls it governance minimization: the deliberate reduction of how much a protocol relies on, and how much power it hands to, active governance. “Minimizing governance calls for protocols to reduce their reliance on, and the power of, decentralized governance,” he wrote, warning that protocols carrying needless governance “would eventually succumb to competition given the costs imposed,” in his essay on Machiavelli for DAOs. The lending protocol Liquity took the idea to its logical end and eliminated human governance altogether; there is nothing to vote on, so there is nothing to capture.

Read through the lens of the a16z attack equation, minimization is the cleanest defense of all. If the vote controls no parameters, holds no treasury and can upgrade nothing, then the value of the attack in the equation is zero, and no amount of cheap voting power makes the raid worth running. You cannot flash-loan a vote that runs nothing.

The mood among the people who funded the DAO experiment has curdled accordingly. “We spent the last 10 years rediscovering the hard way that direct democracy is a bad idea,” a16z crypto general partner Ali Yahya told Fortune. CoinFund’s Jake Brukhman, in the same piece, said on-chain governance “proved to be almost everywhere else but the [block]chain.” Even a defender like Botto’s Simon Hudson conceded the damage: “Are they dead? Certainly not. Have a lot of people been burned by them? Oh yeah.”

The Legal Wrappers: DUNA, DUNI and the Marshall Islands

If the vote cannot be abandoned outright, the next best thing is to put a legal shell around it. By 2026 the entityless DAO, the pure on-chain collective with no company behind it, was going out of fashion fast.

Uniswap led the way. In September 2025 its governance adopted a Wyoming Decentralized Unincorporated Nonprofit Association, branded DUNI, with 99.9 percent support and more than 125 million UNI in favor against just 742 opposed, per the Uniswap governance forum. The wrapper gives the DAO legal standing to sign contracts, pay taxes and retain service providers, and, crucially, caps the personal liability of the people who vote.

Wyoming created the DUNA structure, effective 1 July 2024 for associations of at least 100 members, specifically to give DAOs legal personality and limited liability; a16z called the state an “oasis” for the model when CoinDesk covered the new law. The Marshall Islands offers a competing option, the DAO LLC under its 2022 DAO Act, which grants express legal personality with no corporate or capital-gains tax, as the islands’ registrar sets out. Jennings has described the DUNA as the first legal structure that lets communities govern the services they use without exposing members to unlimited liability. The pitch is precisely the fear the Sky episode made concrete: without a wrapper, a DAO’s members can be treated as partners in an unincorporated association, jointly on the hook for whatever the collective does.

WrapperJurisdictionLegal personalityMember liabilityNotable adopter
DUNA (nonprofit association)Wyoming, USYesLimitedUniswap (DUNI)
DAO LLCMarshall IslandsYesLimitedVarious offshore DAOs
FoundationCayman, Switzerland, othersYesLimitedENS, ApeCo
No wrapper (default)NoneNoUnlimited, joint and severalOoki DAO, by court ruling

When the DAO Becomes a Person: Ooki, Sky and Sanctions Law

The liability fear is not hypothetical, and the precedent is three years old. In June 2023 the Commodity Futures Trading Commission won a default judgment against the Ooki DAO, with a federal court holding that the DAO was an “unincorporated association” and a “person” under the Commodity Exchange Act, liable for a $643,542 penalty, and ordering its website shut down, the CFTC announced. The holding meant the token holders who voted were members of a liable association, not anonymous participants in a diffuse experiment.

The Sky scare is the same logic applied to sanctions rather than derivatives law. If MKR and SKY holders can vote to reconfigure the peg stability module, they are the people with the power to stop illicit flows, and so, in the reading of analysts who studied the Inca Digital findings, they are the people a regulator could pursue for failing to. The report pointed to concentrated governance tokens, administrative privileges and control over upgrades as the markers of a system that still has identifiable controllers, however decentralized it calls itself. The same question of who answers when an autonomous system misbehaves is playing out across crypto, from stablecoins to on-chain AI agents.

It also lands on a nerve the industry has spent the year poking at. A decentralized protocol has no KYC desk and no compliance officer, which reads as a feature until sanctioned money flows through and there is no one to subpoena except the voters. Crypto’s wider blind spots on illicit finance make the governance layer an obvious place for enforcement to start. The law here is genuinely unsettled, which cuts both ways: when Avraham Eisenberg was prosecuted over his $110 million Mango Markets exploit, a judge vacated his convictions in May 2025, reasoning in part that Mango’s contracts were permissionless and did exactly what they were coded to do, CoinDesk reported. If using a governance system as designed is not obviously a crime for the attacker, the responsibility may land instead on the people who built and still steer the system.

Washington Draws a Line: the SEC’s New FAQ

The retreat into foundations and wrappers raises an awkward question for American regulators: if a DAO hands control to a small board, is it still decentralized, or has it just admitted it had controllers all along? On 2 November 2026, the SEC’s Division of Corporation Finance tried to answer, adding three questions to its crypto-asset FAQ covering governance, control and decentralization, as law firm Lowenstein Sandler summarized.

The staff drew a careful line. Participants in a broad-based governance system “generally will not be considered a group constituting a central party” when they are unaffiliated and have no agreement, formal or informal, to act together, the FAQ says. But a central party exists where someone holds the authority to control access at the protocol, validator or user level, including the unilateral power to restrict access, set governance processes or change the rules. Decentralization, the staff stressed, means a system runs autonomously, “without a person or group exercising operational, economic, or voting control,” and it is distinct from mere functionality.

Concentration of operational, economic or voting authority, in other words, is exactly what pulls a project back inside the securities perimeter, and a foundation is fine only so long as it does not become the controller. That is a fine needle to thread. A protocol minimizes its governance to avoid being captured or sued, then has to prove the resulting structure is not so centralized that its token looks like a security. The guidance is only staff views, not Commission rules, so it can shift. For now it tells founders what Washington will be looking at: who actually holds the keys, the money and the votes.

For founders the message is uncomfortable but clear. A wrapper or a foundation can cap liability and give a protocol someone to answer the phone, yet the same structure that makes a DAO legible to a court can make its token look less like a commodity and more like a security if the new center holds too much sway. The safe harbor is narrow: enough structure to be accountable and compliant, not so much control that the project stops being decentralized in the staff’s sense. Most of 2026’s retreats were bets that this needle can be threaded.

The Retreat Is Also the Defense

Strip away the legal and philosophical framing and the retreat is, at bottom, a security measure. Every incident in the 2026 attack wave followed the same four steps, buy the votes, write the proposal, win the ballot, drain the treasury, and every defense that works does so by breaking one of them. Dissolving the DAO removes the treasury. Minimizing governance removes the proposal’s power. A foundation or a security council removes the attacker’s ability to execute unilaterally. A monitoring firm, Blockaid, counted at least seven governance takeovers between June and August that drained about $22 million across Ethereum, Solana and Base, and its remedies read like a blueprint for the retreat: raise quorum, ring-fence the treasury from any single vote, and put an empowered human veto on the execution path, as the firm detailed in its post-mortem of the wave.

TargetDateLossHow the vote was seized
BonkDAOJuly 2026About $20 millionSeven wallets controlled nearly the entire vote
Term FinanceAugust 2026About $8.5 million (68% of vaults)Cheap token bought past quorum; timelock elapsed
X100October 2026About $4.85 millionAave flash loan seized two of three Senate seats in one transaction
Marinade (thwarted)September 2026$0DAO council vetoed two malicious proposals within six hours

Marinade’s near miss is the proof of concept for the new doctrine. When an attacker exploited a voting bug to file two malicious proposals in September, Marinade’s DAO council rejected both within six hours, roughly four days before they could take effect, and no funds moved, KuCoin reported. The brake worked. But it worked because a small, trusted committee could override the vote, which is to say it worked by being a little less decentralized.

That is the catch that runs through the whole retreat. Every brake re-centralizes, and every new center becomes the new target. The security council that can veto a malicious proposal can also be captured, coerced or subpoenaed. The foundation that can sign a contract can also be served one. Protocols are not escaping the problem so much as relocating it, from a diffuse crowd that is hard to sue to a concrete body that is easy to find, which is better for users and considerably worse for the handful of people who now sit on that body.

The relocation is not only legal. A guardian holding the keys to pause a protocol is a richer target for the very attackers it exists to stop, and a cleaner one for a subpoena. Term Finance had a seven-day timelock and veto rights and still lost most of its vaults, because the brake meant nothing with no one standing behind it in time. A defense that depends on a small group being awake, honest and fast is a defense that has quietly reintroduced the single point of failure that decentralization was supposed to remove.

What the Retreat Leaves Token Holders

For anyone holding a governance token, 2026 rewrote what that token is. It used to be sold as a share of control. Increasingly it is a backstop right at best: a claim on a treasury during a wind-down, or a seat in a system whose real decisions happen in a boardroom. None of that is necessarily bad, because a professionally run foundation may serve users better than an apathetic townhall ever did. But it is different from the promise, and holders should price the difference. Before treating a token vote as meaningful control, it is worth checking what it still governs.

  • What does the vote actually control now: live parameters and a real treasury, or only symbolic signaling while a foundation or core team holds the keys?
  • Is there a legal wrapper, and who does it protect? A DUNA or foundation can cap your liability, but it also moves real authority to directors you did not elect directly.
  • Who can move the treasury? If any single proposal, council or multisig can reach the funds, that is both the attack surface and the point of legal control.
  • Is there an empowered veto, and who sits on it? A security council is only as good as its members and their willingness to act in time, as Term Finance learned when its unused brake simply elapsed.
  • How concentrated is the vote? If the top handful of wallets can pass anything, your vote is decorative and the token is cheap for a raider to seize.

So is the DAO dead? The honest answer is the one Botto’s Simon Hudson gave Fortune: not dead, but bruised, and plenty of people got burned. The pure, leaderless, one-token-one-vote organization that was supposed to replace the company turned out to be too easy to capture and too hard to defend, both from raiders who wanted the treasury and from regulators who wanted a defendant. What is replacing it is more honest about power: foundations that admit they are in charge, wrappers that admit someone can be sued, minimized governance that admits most decisions should never go to a vote at all.

The retreat is, in the end, a concession that the vote was never the point. The point was the protocol, and in 2026 crypto decided the protocol was worth protecting, even from its own voters. The leaderless dream is not gone, but it has learned to keep a boss on call, a lawyer on retainer and a brake within reach, which is a long way from where it started.

Frequently Asked Questions

What is a governance attack?

A governance attack is when someone acquires enough of a protocol’s voting power to pass a malicious proposal and redirect its treasury or permissions, without breaking any code. The smart contracts execute exactly as written, so the exploit is the vote itself. Documented 2026 cases include BonkDAO, which lost about $20 million, and X100, which lost about $4.85 million after an attacker used an Aave flash loan to seize control.

Why are DAOs moving to foundations and legal wrappers in 2026?

Token-holder governance turned out to be both an attack surface, because it is cheap to seize when tokens are thin and turnout is low, and a legal-liability surface, because a DAO can be treated as a liable person, as in the CFTC’s Ooki case. Foundations, Wyoming DUNAs and Marshall Islands LLCs cap member liability and give the organization legal standing, which is why protocols like ENS and Uniswap adopted them.

Can a DAO be held legally liable for how its protocol is used?

It can. In 2023 the CFTC won a default judgment treating the Ooki DAO as an unincorporated association and a person under the Commodity Exchange Act, making token holders who voted potentially liable. In November 2026, analysts warned that Sky’s token holders, who can halt or reconfigure its peg stability module, could face sanctions exposure after $518 million in illicit flows were traced through it.

What is governance minimization?

Governance minimization is the design philosophy, advocated by a16z general counsel Miles Jennings, of reducing how much a protocol relies on active governance, ideally toward zero. Liquity, for example, eliminated human governance entirely. Less governance means a smaller attack surface and less for a raider to seize, which is why it became a default for new protocols in 2026.

Does the governance retreat mean DAOs are dead?

Not quite. Investors like a16z’s Ali Yahya argue the last decade proved direct democracy is a bad idea, and most large protocols have handed day-to-day power to foundations, councils or founders. But token holders usually keep ultimate backstop rights, and the vote still exists; it just controls far less than the original leaderless-organization dream promised.

By Anneke de Vries, DAO governance and security correspondent at HOGE Wire.

Share 𝕏 Post Telegram