Blockchain Governance and DAOs

Table of Contents

    Blockchain governance determines who can change a protocol’s rules, spend its treasury or replace its software. For an investor or user, those permissions can matter as much as the rules operating today: a lending market’s risk settings or a treasury’s spending policy may change after you commit funds.

    A decentralized autonomous organization, or DAO, is one way to organize those decisions. To evaluate one, follow a proposal from discussion to execution, identify the people or contracts with authority at each step, and check what token ownership actually gives you. The examples below separate technical control from legal rights. Legal discussion uses U.S. sources, with Wyoming entity-law examples, checked in November 2026.

    What Is DAO Governance?

    A DAO coordinates participants through shared governance rules, often using blockchain tokens and smart contracts. Its members might decide on grants, treasury spending, protocol parameters or software upgrades. The label does not establish how decentralized the organization is, whether every action is automated, or which legal entity owns its assets.

    Three Rights to Trace Separately

    Proposal rights determine who can put an action to a decision. Voting rights determine how support is counted. Execution authority determines who or what can actually change code, move funds or enter an agreement. These rights can belong to different actors.

    OpenZeppelin’s Governor framework makes voting power, counting rules and execution arrangements configurable. That is a useful starting point: there is no single DAO voting system to assume from a token’s name. [1] Our smart-contract guide explains the underlying programs; this article focuses on who can change or direct them.

    Governance versus blockchain consensus

    Consensus determines which transactions and blocks a network accepts under its existing rules. Governance concerns decisions about rules, permissions and resources. A DAO treasury vote can run as an application on Ethereum without giving its voters authority to change Ethereum’s consensus rules. Likewise, holding an application’s governance token does not automatically make someone a validator. See proof-of-stake consensus for that separate role.

    How a Proposal Moves From Discussion to Execution

    Consider a proposed treasury grant. The description says which team should receive funding, but an executable proposal also needs the actual recipient, asset and amount. A typical process has the following stages, although a particular organization may combine or omit some of them:

    1. Discussion and screening. Participants debate the idea, conflicts, costs and alternatives. A forum poll may only measure sentiment.
    2. Formal submission. An eligible proposer submits the decision and, where applicable, the transaction payload. A proposal threshold can require a minimum amount of voting power before submission.
    3. Voting-power snapshot. The system determines eligible weight at a specified block or time. Check whether delegation or token locking must occur before that point.
    4. Voting and counting. The voting period closes. Quorum and approval conditions are checked separately, including how abstentions count.
    5. Queue and review delay. A successful proposal may enter a timelock, a required wait before execution. Cancellation or emergency powers may still apply.
    6. Execution and verification. A permitted transaction applies the approved calls, or authorized signers implement the decision. Verify the resulting transfer, parameter or contract implementation.

    An on-chain result marked succeeded is therefore not necessarily an executed change. OpenZeppelin documents distinct proposal, queue and execute steps. Its timelock also has proposer, executor, canceller and administrator roles that affect the path. [1]

    Compound provides a concrete governance example. Its v2 documentation separates the COMP token, Governor Bravo and Timelock. Token holders delegate votes; the governance contract evaluates proposals; the Timelock queues and executes approved actions. This architecture illustrates the different jobs without implying that the documentation’s published parameter values are a substitute for checking a particular deployed contract. [2]

    On-Chain vs Off-Chain Governance

    The location of the ballot and the mechanism that implements its result are separate design choices. A signed off-chain ballot can be connected to on-chain execution, while an on-chain ballot may still leave important assets or decisions outside its scope.

    How the voting result reaches an action
    Arrangement Where the decision is recorded Execution question
    On-chain Governor Votes and proposal state are recorded through contracts. Does the Governor or timelock actually hold the required authority over the target?
    Off-chain ballot with manual execution Participants sign ballots outside the chain’s transaction process. Which signers must implement the result, and what prevents refusal or a different transaction?
    Off-chain ballot with an execution integration The ballot is off-chain; another mechanism connects the result to a contract. What oracle, challenge process, module or other condition authorizes execution?

    Classic Snapshot offers off-chain voting without a gas fee for casting the ballot. Its SafeSnap integration connects governance to a Safe treasury using an oracle, and Snapshot X is a separate fully on-chain voting protocol. Calling everything bearing the Snapshot name an advisory poll would miss these differences. [3]

    Compare the full arrangement rather than choosing solely on transaction costs. Manual execution introduces signer discretion; automated execution depends on correctly specified permissions and code. Neither arrangement establishes legal authority over a bank account, company or intellectual property merely by recording a vote.

    How Delegated Voting Changes Power

    Delegation allows a holder to assign voting power to a representative under the token’s rules. In Compound’s documented design, a holder can delegate to another address or to themselves. Delegation changes the votes available to the representative without itself transferring the underlying COMP tokens. The representative’s voting weight and the holder’s token balance therefore answer different questions. [2]

    Delegates can reduce the time individual holders spend reviewing proposals. They can also concentrate decisions among a small set of active representatives. Evaluate the delegate’s expertise, voting record, compensation, conflicts and explanation of decisions. A large delegated balance does not tell you whether the delegate owns those tokens or must bear their full economic losses.

    Timing matters. If a proposal uses a past snapshot, a later purchase or redelegation may affect future proposals without changing weight for the current one. Read the specific voting-token and Governor rules before treating a current wallet balance as usable voting power.

    This is also an asymmetric-information and principal-agent problem: holders may rely on delegates or developers who have more information and different incentives. Public vote records help, but they do not disclose every private relationship.

    DAO Quorum and Approval: A Worked Example

    Quorum is a minimum qualifying vote weight required for a decision to proceed. Approval asks whether the proposal received sufficient support under a separate rule. Quorum may count only For votes, For plus Abstain, or another specified set; do not infer the convention from a participation percentage.

    For example, OpenZeppelin’s illustrated GovernorCountingSimple counts For and Abstain toward quorum, while Compound’s v2 documentation describes a minimum number of votes in support. These are different rules. [1] [2]

    Now consider a fictional DAO with 1,000,000 eligible voting units at a fixed snapshot, one vote per unit, and 300,000 units delegated. Assume formal submission requires 20,000 delegated units. For a submitted proposal, quorum requires For plus Abstain equal to at least 20% of eligible supply, and approval requires strictly more For than Against. These invented rules are not Compound’s parameters.

    Hypothetical ballots after voting closes
    Ballot choice Voting units Counts toward this quorum?
    For 150,000 Yes
    Against 60,000 No
    Abstain 30,000 Yes
    All submitted ballots 240,000 Includes all three choices
    A Majority Can Still Fail Quorum

    Approval share among decisive votes: 150,000 / (150,000 + 60,000) × 100 = 71.43%.

    Overall turnout: 240,000 / 1,000,000 × 100 = 24%.

    Required quorum weight: 20% × 1,000,000 = 200,000.

    Qualifying weight: 150,000 + 30,000 = 180,000, or 18% of eligible supply.

    The proposal has more For than Against but fails quorum by 20,000 units. Total turnout above 20% does not fix that failure because Against votes do not qualify under the assumed rule. With the same ballots and a different rule counting all choices, quorum would pass. Abstention is consequently different from both an Against vote and failing to vote.

    Also label the approval denominator: For represents 150,000 / 240,000 = 62.5% of all submitted ballot weight, a different statistic from 71.43% of decisive votes. The 20,000-unit submission threshold answers who can propose; it does not replace either success test.

    How to Measure Governance Concentration

    Use a dated snapshot and distinguish eligible supply, delegated voting power and votes actually cast. A percentage without its denominator can make the same delegate look small or dominant.

    Delegate Concentration
    Top-k delegated share = Voting power of the k largest delegates / Total delegated voting power × 100
    A concentration measure for the defined snapshot and delegation system, not a complete measure of independence or control.

    Use a separate hypothetical concentration scenario, keeping the totals at 1,000,000 eligible units and 300,000 delegated units. Suppose the three largest delegate addresses have 135,000, 90,000 and 45,000 units, and ten other delegate addresses have 3,000 each. The distribution totals 135,000 + 90,000 + 45,000 + (10 × 3,000) = 300,000. This scenario does not reuse the earlier ballot tally.

    Hypothetical concentration depends on the denominator
    Measure Calculation Result
    Largest delegate / eligible supply 135,000 / 1,000,000 13.5%
    Largest delegate / delegated power 135,000 / 300,000 45%
    Top three / delegated power 270,000 / 300,000 90%

    The largest address controls less than one-seventh of eligible supply but nearly half of delegated power. Another useful measure is the smallest number of delegates whose combined weight exceeds a specified threshold. Here, one has 135,000 units, below half of 300,000; the top two have 135,000 + 90,000 = 225,000, above half. Two of the thirteen addresses can therefore command a majority of delegated power.

    For a separate possible ballot outcome, suppose those two vote all 225,000 units For and the remaining 75,000 delegated units vote Against. They would satisfy both the earlier 200,000-unit quorum and For-greater-than-Against rule with 225,000 / 1,000,000 = 22.5% of eligible supply. This scenario illustrates potential control; it is not the earlier ballot tally.

    Address counts have limits. One organization may control multiple addresses, while a single delegate may represent many unrelated holders. Report known common control and uncertainty. Do not add delegated power to the underlying token holdings: they can represent the same units twice. Broader altcoin analysis also needs issuance, vesting and liquidity checks that these voting metrics do not supply.

    Execution Risks: Multisigs, Upgrades and Governance Attacks

    A governance dashboard may show an open vote while a separate address can upgrade the treasury contract. Trace permissions on the actual assets and contracts, including these control points:

    • Treasury signers: who can authorize a transfer, change owners or change the signature threshold?
    • Upgrade administrator: who can replace the code behind an upgradeable contract?
    • Timelock roles: who can queue, execute or cancel actions, and who can change those roles?
    • Emergency powers: can an actor only pause specified functions, or also move assets or bypass the normal process?
    • Execution modules: does another permitted route execute transactions outside the ordinary signer path?

    For example, a hypothetical 3-of-5 multisig requires three of five owner addresses for its ordinary signed transaction path. It does not prove five independent people or that those owners are bound to follow a token poll. Safe’s documentation also describes module transactions that bypass the ordinary signature-verification path. A threshold label alone is therefore incomplete without the account’s actual configuration. [4]

    Similarly, a timelock helps only within the permissions it governs. Extra proposal or administrator powers can create alternative routes. A waiting period gives users time to investigate or react, but does not promise that they can withdraw, sell or obtain a favorable price before execution. [1]

    Beanstalk: temporary power and a historical exploit

    Beanstalk Farms reported that an attacker used a flash loan to exploit Beanstalk’s governance mechanism on April 17, 2022 and transfer funds to an attacker-controlled wallet. This historical event concerns the design operating at that time, not a claim about Beanstalk’s present configuration. [5] In a same-transaction flash loan, borrowed assets must be returned with any required fee before the transaction ends, or the transaction reverts. Some lending products also support conversion to a properly backed debt position, so check the actual arrangement. [9]

    The evaluation lesson is to ask how long an attacker must control qualifying power, when it is measured, and what an approved action can reach. Past snapshots, voting delays and limits on emergency execution can address particular attack paths, but their effectiveness depends on implementation and the rest of the system. They are not interchangeable guarantees.

    Capture can also involve purchased or delegated voting power, coordinated insiders, compromised keys or a misleading description paired with harmful executable calls. A token’s market capitalization is not the cost of acquiring control: liquidity, delegation, borrowing, threshold rules and the value accessible through execution all matter. Treat code review, key management and incident response as part of operational risk management, alongside ballot analysis.

    A token’s ability to vote on a contract does not by itself prove membership in a company, ownership of a bank account, a right to profit distributions or authority to sign an employment agreement. Trace each claimed right to the relevant entity, governing documents, contracts and applicable law. Technical execution and legal enforceability are different questions.

    A legal wrapper is an entity structure used to connect an organization with legal rights and obligations. It may hold property or contract with service providers, but it matters which assets and activities it actually covers. A foundation or service company associated with a protocol is not automatically owned or controlled by every token holder.

    Wyoming illustrates why the precise form matters. Its DAO LLC supplement and decentralized unincorporated nonprofit association, or DUNA, statute provide different arrangements. The following is a state-specific comparison of the official compilation retrieved in November 2026. [6]

    Two Wyoming structures have different conditions
    Structure Selected statutory requirements Question for a participant
    Wyoming DAO LLC An LLC elects DAO status through its articles. The supplement addresses smart-contract identifiers and governance, and requires utilized smart contracts to be upgradeable. How do the articles, smart contracts and any operating agreement define membership and voting rights, including their statutory priority rules?
    Wyoming DUNA At least 100 mutually consenting members, a common nonprofit purpose and election under the act. Profit-making activities must support that purpose; distributions are limited, with specified exceptions. Does the organization qualify, and are any compensation, benefit or repurchase arrangements permitted by the statute and governing principles?

    These structures include statutory protections against liability based solely on member or manager status, with the details differing by form. They do not make unlawful conduct permissible or eliminate every basis for individual liability. A token project does not acquire either structure merely by calling itself a DAO. Other states, countries and federal laws require their own analysis. [6]

    Ooki DAO is a bounded enforcement example. In June 2023, a federal district court granted default judgment in the CFTC’s case and treated Ooki DAO as an unincorporated association subject to suit under the Commodity Exchange Act. The court evaluated the particular pleaded facts in a default proceeding. This was a judgment against the DAO; it does not establish that every holder or voter in every DAO is personally liable. [7]

    Securities treatment is another separate question. The SEC’s March 2026 interpretation distinguishes a crypto asset’s characteristics from circumstances in which its offer or sale is subject to an investment contract, and recognizes that certain digital commodities can carry governance rights. Neither the word governance nor a claim of decentralization supplies a classification for an individual offering. Broader obligations belong in a jurisdiction-specific review; see our cryptocurrency regulation guide for context. [8]

    How to Evaluate a DAO Before Participating

    1. Identify what is governed. List the treasury assets, contracts and off-chain entities. Separate powers the token controls from powers retained elsewhere.
    2. Read the actual decision rules. Record proposal eligibility, snapshot timing, quorum numerator and denominator, approval threshold and abstention treatment.
    3. Measure power at the right time. Compare holder, delegate and actual-voter concentration; identify known affiliated addresses and upcoming changes in eligible supply.
    4. Trace every execution path. Identify the Governor, timelock, owners, upgrade administrators, modules and emergency roles. Check who can change those permissions.
    5. Match description to action. Have executable calls and recipient addresses checked. Confirm a completed proposal changed the intended state rather than stopping at a successful ballot.
    6. Check incentives and legal exposure. Review delegate compensation, related-party grants, entity documents, asset title, participant roles and the jurisdictions involved.

    Keep this record dated. A new delegation, module, administrator or governance amendment can change the conclusion even if the token’s name stays the same.

    Common Mistakes About DAO Governance

    • Equating one token with one person. Token-weighted voting counts units; identities and independence require separate evidence.
    • Using wallet ownership as the voting-power ranking. Delegation and snapshot rules can produce a different ranking.
    • Calling every majority a passed proposal. Check quorum, approval, cancellation and execution separately.
    • Treating abstention as nonparticipation. It may count toward quorum even when it does not count toward approval.
    • Trusting a multisig label without permissions. Owner affiliation, modules and upgrade authority can change the practical control picture.
    • Assuming a token vote grants legal ownership or immunity. Determine the actual rights, entity and applicable law.

    Limitations of Governance Analysis

    An Observable Vote Is Only Part of Control

    A transparent ballot does not reveal every beneficial owner, private agreement, compromised key or code vulnerability. Concentration statistics and published rules help identify questions; they do not certify a protocol’s safety, a token’s value or a participant’s legal position.

    Low participation can increase an active coalition’s influence, but a low-turnout proposal is not automatically malicious. High quorum can discourage capture while making legitimate changes harder to pass. Emergency powers may limit an unfolding incident while concentrating discretion. Assess these tradeoffs against what the system can change and what participants stand to lose.

    Finally, this guide is not a protocol audit or a comprehensive legal survey. The numerical examples are hypothetical; named systems illustrate specific documented features or historical events. Verify current contracts, governing documents and legal requirements before acting.

    Frequently Asked Questions

    Only if you satisfy that system’s submission rules. A protocol may require delegated voting power above a threshold, a deposit or another eligibility condition. Forum participation, proposal submission and voting can have different requirements.

    Yes, depending on its quorum and approval rules. Many token-weighted systems test vote weight rather than the number of people or wallets. Always state whether a percentage refers to eligible supply, delegated power, all submitted ballots or only decisive For and Against votes.

    Not in Compound’s documented delegation mechanism: it assigns voting power without transferring the underlying token ownership. Other arrangements may involve locking, custody or separate contracts. Check the actual transaction and permissions rather than assuming every product using the word delegation works the same way.

    The ballot alone does not establish that result. Authorized signers may need to make a transaction, or a configured integration may connect the result to contract execution. Identify the mechanism, its conditions and the account permissions that actually control the funds.

    No universal immunity follows. Wyoming’s LLC framework distinguishes company obligations from liability based solely on member or manager status, but statutory exceptions and other applicable laws still matter. The entity, conduct, legal documents and jurisdictions need individual analysis.

    No. It delays actions on the paths it governs. Its protection depends on who can alter roles, whether another path bypasses the delay, what monitoring occurs and which responses remain possible. Voting-power measurement, code correctness, signer security and emergency permissions also matter.

    Sources

    1. OpenZeppelin: How to set up on-chain governance. Configurable Governor rules, snapshots, timelock roles and proposal lifecycle.
    2. Compound: v2 Governance, first-party documentation source. Delegation, Governor Bravo, quorum and Timelock. Documentation architecture is not a live verification of deployed parameters.
    3. Snapshot: Getting started, SafeSnap and Snapshot X overview. Different ballot and execution arrangements.
    4. Safe: Smart Account Concepts. Owners, thresholds, ordinary transactions and module transactions.
    5. Beanstalk Farms: Beanstalk Governance Exploit, April 19, 2022. First-party account of the April 17 incident.
    6. Wyoming Legislature: Title 17. Sections 17-29-304; 17-31-104 through 109 and 115; 17-32-102 through 109. Current official compilation retrieved November 2026.
    7. CFTC v. Ooki DAO: Order Granting Motion for Default Judgment, N.D. California, June 8, 2023, especially pages 1 and 7–8. Historical entity-level ruling under the Commodity Exchange Act.
    8. SEC: Application of the Federal Securities Laws to Certain Types of Crypto Assets and Certain Transactions Involving Crypto Assets, March 17, 2026, effective March 23, 2026, sections III.A, III.E and IV.A–B.
    9. Aave: V3 Flash Loans. Same-transaction repayment/reversion and the distinct supported debt-opening option.

    Sources accessed November 15, 2026. Governance parameters and legal interpretations can change. All voting distributions and numerical calculations in this article are hypothetical.

    Disclaimer

    This article is for educational and informational purposes only and does not constitute investment, legal, tax or cybersecurity advice. Named protocols and historical cases are examples, not recommendations or security assessments. Token and protocol participation can involve substantial loss and jurisdiction-specific obligations. Consult qualified professionals about your circumstances.