Ooki is the case everyone reaches for, and its reach is narrower than the shorthand. It was entered by default – the DAO never appeared to defend itself – and what it decides is that the organization is a person capable of being sued and fined. Whether an individual member must answer for that fine is a different question, and it has been litigated in two other cases which disagree with each other on exactly the point a one-line summary flattens.
In Sarcuni v. bZx DAO, 664 F. Supp. 3d 1100 (S.D. Cal. 2023), the court accepted an allegation of a general partnership comprising every holder of the DAO's token. In Samuels v. Lido DAO, No. 3:23-cv-06492-VC (N.D. Cal.), Judge Vince Chhabria's order of 18 November 2024 held that Lido DAO was adequately alleged to be “a general partnership within the meaning of California law” – California treats an association carrying on a business for profit as a partnership “whether or not the persons intend to form a partnership” (Cal. Corp. Code § 16202(a)) – and that three of the four institutional investors sued (Paradigm Operations, Andreessen Horowitz and Dragonfly Digital Management) were adequately alleged to be partners and “therefore liable for Lido's conduct”; the fourth, Robot Ventures, was dismissed.
But the same order expressly distinguished Sarcuni and drew the line differently for retail holders: “every LDO holder, on the other hand, hasn't automatically joined the partnership … because they don't all necessarily have the ability to meaningfully participate in DAO governance.” On that reading it is the capacity to participate in governance, not the holding of a token, that puts someone in the frame.
Both rulings are pleading-stage: each decided only that the claim could proceed, and the existence of a partnership is a question of fact reserved for a full record. Neither has produced a judgment against a token holder. Sarcuni's docket closed on 16 November 2023 without one. Samuels is unresolved and now on appeal – a notice was docketed in the Ninth Circuit as Samuels v. AH Capital Management, No. 25-5701 on 9 September 2025, and the district court stayed proceedings pending that appeal with joint status reports every 120 days, the most recent filed on 18 June 2026. Read on 7 September 2026, neither docket is terminated.
The defensible statement of the default risk is therefore narrower than the shorthand: two federal district courts have held the general-partnership theory plausible enough to litigate, they disagree about how far down the holder base it reaches, and no court has decided it on the merits. A wrapper is bought against that uncertainty rather than against a settled rule – which is also why the forms below are drafted to make the question moot rather than to win it.
A DAO legal structure (or "legal wrapper") is a recognized legal entity that a DAO adopts so it can exist in the eyes of the law: sign contracts, hold off-chain assets, pay taxes, appear in court – and, above all, shield its members from personal liability. The code that runs a DAO on-chain says nothing about how a court will treat the humans behind it, and the gap between those two worlds is where wrappers live.
A DAO that never incorporates anywhere is, in most jurisdictions, likely to be treated as a general partnership or unincorporated association – meaning individual members can be personally liable for the organization's obligations. The landmark test was CFTC v. Ooki DAO: in June 2023 a U.S. federal court entered a default judgment holding that the Ooki DAO was a “person” under the Commodity Exchange Act, could be held liable for operating an illegal trading platform, and owed a $643,542 penalty – a precedent the CFTC itself called precedent-setting. The founders had adopted the DAO structure explicitly hoping to avoid legal accountability; the court disagreed. That risk is what drives the entire legal-wrapper workstream (see the legal-recognition arc in the history of DAOs).
Wyoming's DAO Supplement (passed March 2021, effective July 2021) was the first U.S. statute to let a DAO register as a limited-liability company. Members get the standard LLC shield and pass-through taxation, and formation is fast. The trade-offs: the filing must identify the DAO's smart contracts, the LLC must show activity on a rolling one-year basis or face dissolution, and U.S. formation brings registered-agent and reporting obligations that make full member anonymity impractical. Commentators generally recommend it for governance-focused DAOs rather than investment DAOs, which trip U.S. securities rules.
The DUNA did not appear from nowhere. It descends from A Legal Framework for Decentralized Autonomous Organizations, the 2021 paper by David Kerr (Cowrie) and Miles Jennings (a16z crypto) that first proposed the unincorporated nonprofit association (UNA) as a domestic home for DAOs. Its starting observation is that DAOs fit none of the existing categories cleanly: they are "analogous to partnerships, and yet not partnerships; analogous to corporations; and yet not corporations."
The paper's threshold question is whether a DAO's purpose is for-profit or not-for-profit, and it stresses that "not-for-profit purpose is not equivalent to tax-exempt" – few DAOs will qualify as tax-exempt, but many meet the requirements of a not-for-profit for state-law purposes, and those should consider registering as a UNA in a state that recognises the form. From there it sets out the two structural options that later legislation drew on: wrapping an entire DAO as a single UNA, or siloing it – the treasury wrapped in a UNA, with the protocol left regimeless or placed in a separate vehicle, so that grants, contributor pay, and treasury diversification sit under one entity while protocol parameters sit under another.
The paper is also candid that this is a workaround rather than a solution, arguing that there "remains a significant need for federal and state legislators to modify existing legal entity structures" – the need Wyoming's DUNA statute went on to address.
Wyoming followed with the Decentralized Unincorporated Nonprofit Association – SF50, signed March 7, 2024 and effective July 1, 2024. Where the DAO LLC borrows a corporate form, the DUNA was purpose-built for genuinely decentralized memberships: it gives the association legal personhood (it can contract and appear in court), limits each member's liability for the actions of others, and lets the DAO meet tax and reporting obligations – while still permitting profit-generating activity and member compensation despite the "nonprofit" label. a16z crypto, which drafted the model legislation, describes it as legal infrastructure for the network itself rather than a business wrapper around it.
The first sustained working example is the Syndicate Network Collective, a Wyoming DUNA formed in August 2025 to govern Syndicate’s appchain network. It is worth reading as the concrete form of everything above: the collective publishes accrual-basis quarterly financial statements with a tax position, elected U.S. C-corporation treatment at the 21% federal rate, and states in its member disclosures that owning the token and voting is election to membership – with disputes routed exclusively to the Wyoming Chancery Court and class actions waived. It also shows the wrapper’s limits. Nearly a year after formation its governance venue had recorded no proposals at all, and 1.42% of the token supply was votable. A DUNA supplies the legal apparatus for tokenholder control; it does not supply the tokenholders.
Uniswap adopted the form in September 2025. Its governance became DUNI, a Wyoming DUNA, by an on-chain vote executed on 12 September, which set aside $16.5m of UNI for legal defence and back taxes and made a Wyoming firm, Cowrie, its administrator (Uniswap governance). The proposal states that the entity changes nothing about the protocol, the token or on-chain voting; its job is to let governance sign contracts, retain service providers and pay tax.
The Republic of the Marshall Islands enacted a dedicated DAO Act in 2022, creating an offshore DAO LLC with legal personhood that explicitly recognizes on-chain governance: the smart contract can serve as the operating agreement, no directors are required, and and members can act pseudonymously in public – though not to the state: §712 puts every beneficial owner's passport number and wallet addresses on a report held by the Registrar. It has become the default offshore choice for DAOs that want their on-chain rules to be the legal rules; most obtain it through MIDAO, the RMI’s exclusive registrar.
The Radix network's DAO has chosen this wrapper: its draft certificate of formation names Radix DLT DAO LLC, a non-profit Marshall Islands DAO LLC, with MIDAO as registered agent. On 17 September 2026 the council forming it was still completing its submissions to MIDAO, so nothing had been filed with the registry. radix.wiki tracks the filing on its incorporation card.
The Cayman foundation company predates DAO-specific law but became a crypto workhorse: a company without shareholders, purpose-driven rather than profit-distributing, requiring at least one director and a supervisor who act as the legal layer between community votes and off-chain execution. It is the common choice for protocol foundations and token issuers that need a robust, court-tested vehicle more than on-chain purity. The older European route serves the same role – Swiss foundations and associations, most famously the Ethereum Foundation (a Swiss Stiftung), anchor many of the largest protocols. In October 2026 Aave's token holders voted to form a Cayman foundation company of their own, the Aave Foundation, to hold the protocol's trademark, domains and code, with no seat for the company that builds it.
A BORG, short for cybernetic organisation, applies the Cayman form to something narrower than a whole DAO. MetaLeX Labs, which designed it, describes a conventional entity whose charter binds it to act through smart contracts (MetaLeX), usually a Cayman foundation company with no members and no owners (MetaLeX FAQ). On-chain, the entity's assets sit in a Safe multisig fitted with MetaLeX's borg-core guard, a contract that rejects every transaction the BORG's policy does not allow, down to individual functions and parameters and with minimum delays between actions. Add-on modules handle tasks such as grant payments or removing a signer, and the guard stores links to the entity's legal documents and to the parent DAO's EIP-4824 record (borg-core). MetaLeX lists a Neutron grants BORG and a Lido Alliance foundation among its users.
The result wraps a committee rather than the membership: a grants council or a security council gets an entity and a constrained multisig, while the DAO that created it stays unwrapped. In September 2025 Aragon added BORGs to its app as a plugin, which it describes as a limited-liability wrapper for chosen addresses and roles: the DAO picks the roles, enters a MetaLeX signer address and approves a proposal to deploy, and MetaLeX does the entity's legal work (Aragon).
Before any of these statutes, Nexus Mutual used an ordinary English form: a member of the mutual was a member of a UK company limited by guarantee, which is why every member passed an identity check, until a February 2022 special resolution moved the capital pool to a Panama private interest foundation. The Law Commission of England and Wales reached the opposite conclusion to Wyoming and the Marshall Islands. Its scoping paper of 11 July 2024 found no need for a DAO-specific legal entity, because there was no consensus on what one should look like and because organisational law should stay technology-neutral, though it told the government to keep the question under review (Law Commission).
It suggested considering instead a limited-liability, not-for-profit association with separate legal personality and flexible governance, similar to the US unincorporated nonprofit associations above, and a review of the Companies Act 2006 so that existing companies can use distributed ledgers in their governance (Norton Rose Fulbright). Boardroom read the paper's attention to hybrid arrangements, an on-chain DAO wrapped in an off-chain entity, as a description of the model most DAOs were already adopting (Boardroom).
The decision usually reduces to three questions. Who must be identified? U.S. forms trade anonymity for domestic legitimacy; the RMI preserves member pseudonymity; Cayman concentrates identity in directors. Where does control really sit? LLC and DUNA forms keep members in charge; a foundation inserts fiduciaries between votes and execution. What does the DAO do? Pure governance travels light; anything touching investors, revenue, or regulated activity needs jurisdiction-specific advice.
Regulators are now engaging with the form directly – on 12 June 2026 Malta's financial regulator published a discussion paper on decentralised finance (Ref 03-2026) covering software-based organisational models and how DeFi interacts with the EU's MiCA regime. Its consultation window closed on 10 July 2026, and as read on 7 September 2026 the MFSA has published no feedback statement, so nothing has yet followed from it – a signal that DAO-specific rules are being considered in the EU, not that any have arrived.
Every wrapper above is written on the assumption that a DAO's decisions arrive through a governing document. Both DAO-LLC statutes then go further than most readers expect: they make the smart contract one of those governing documents, and rank it above the paperwork. Wyoming's DAO Supplement is explicit – under W.S. 17-31-115, "[w]here the underlying articles of organization and smart contract are in conflict, the smart contract shall preempt any conflicting provisions of the articles of organization," with a narrow carve-out for the DAO election itself (17-31-104) and the requirement that the articles publish the contract's identifier (17-31-106(a)–(b)).
W.S. 17-31-109 vests management "in its members, if member managed, or the smart contract, if algorithmically managed," and 17-31-108 demotes the operating agreement to a gap-filler that applies only "[t]o the extent the articles of organization or smart contract do not otherwise provide." The Marshall Islands chapter is drafted the same way: 52 MIRC Ch. 7 §706(3) makes "the certificate of formation or limited liability company agreement or smart contracts" govern relations among members, voting rights, transferability, distributions, and even "[p]rocedures for amending, updating, editing or changing applicable smart contracts," and §708 mirrors Wyoming's management rule.
That ordering matters because the controls a lawyer actually reaches for are ex ante and live in the lower-ranked document: a named multisig signer set, thresholds above which a transfer needs a human, an entrenched asset lock, a duty to refuse execution that would be unlawful, an arbitration clause covering questions of governance validity. An automated settlement layer answers a different question.
Optimistic governance in particular inverts the default – a proposal is treated as approved unless someone vetoes it inside the challenge window – so the on-chain record can reach "approved, tallied, executable" through inaction, while the wrapper's position is that nothing is authorised until each gate has been passed. Where the contract then fires the treasury outflow itself, the payment is arithmetically correct and the entity's own procedural conditions were never tested. A timelock narrows the window in which that can happen without narrowing the authority; it buys reaction time, not consent.
The reverse mismatch is sharper still, because some wrapper obligations cannot be discharged on-chain at all. RMI §712 requires a Beneficial Owner Information Report at formation and with every annual report, identifying each beneficial owner by full legal name, date of birth, street address, "unique identifying number from a nonexpired passport," and the "address(es) and blockchain(s) of all wallets associated with the decentralized autonomous organization" that they hold – held by the Registrar for at least five years after dissolution, with willfully false or omitted reporting declared unlawful. No correctly-tallied vote updates that register. Members also have thin duty-based recourse against one another when the two layers diverge: §709 provides that "no member … shall have any fiduciary duty to the organization or any member except that the members shall be subject to the implied contractual covenant of good faith and fair dealing," and Wyoming's 17-31-110 takes the same line.
There is no settled answer yet, and the statutes decline to give one – RMI §710 even writes a default vote-weight rule (governance tokens held divided by the organization's total "at the time of a vote") that the smart contract is free to override.
The practical drafting responses in use are all attempts to keep the two layers from ever reaching different conclusions rather than to adjudicate between them: give the contract no unilateral outflow path, so on-chain approval produces an instruction that named signers must still execute; write the challenge window and its veto holder into the articles rather than only into the contract, so the preemption rule cuts the same way as the compliance gate; or accept algorithmic management honestly and move the compliance controls into the contract itself, where the statute already says they will win. What does not work is drafting an ex ante control into the operating agreement and assuming a court will read it as a limit on code the same statute has told it to prefer.
A caper is a smart-contract construct on Radix – launching one does not create a legal entity, and Caper does not provide legal wrappers; communities that need one make that jurisdiction choice themselves. What the protocol does address is the enforcement half of the problem: rules that a wrapper would leave to courts and fiduciaries – treasury custody, proposal execution, and each member's guaranteed exit right – are enforced by code, so members are protected from each other by the contract even before anyone is protected from the law by a wrapper.