---
title: "CoW DAO"
url: "https://caper.network/wiki/daos/dexs/cow-dao"
updated: 2026-09-03
license: CC-BY-4.0
license_url: "https://creativecommons.org/licenses/by/4.0/"
---

# CoW DAO

|  |  |
| --- | --- |
| **Name** | CoW DAO |
| **Type** | Protocol DAO governing CoW Protocol — an intent-based, MEV-protecting trading protocol (Ethereum) |
| **Origin** | Incubated inside [GnosisDAO](/wiki/daos/networks/gnosisdao); spun out via [GIP-13 (Phase 2)](https://forum.gnosis.io/t/gip-13-phase-2-cowdao-and-cow-token/2735) with a [$23M funding round](https://www.theblock.co/linked/139692/cow-protocol-raises-23-million-and-spins-out-from-gnosis-dao) (January 2022) |
| **Token** | COW — 1,000,000,000 fixed supply, [≤3% max annual inflation](https://docs.cow.fi/governance/token), changeable only once per 365 days |
| **Treasury share** | 44.4% of COW held by the CoW DAO treasury; 10% to [GnosisDAO](/wiki/daos/networks/gnosisdao), 15% team, 10% [community airdrop](https://blog.cow.fi/cowdao-closes-a-23m-investment-round-9b841abcdbbe) |
| **Governance** | [CoWmunity forum](https://forum.cow.fi/) → CoW Improvement Proposals (CIPs) → gas-free [Snapshot](https://snapshot.org/#/cow.eth) vote |
| **Products** | [CoW Swap](https://cow.fi/) (intent-based batch-auction DEX), CoW AMM (MEV Blocker was co-created by the DAO and [sold to Consensys](https://cow.fi/learn/special-mechanisms-group-acquires-mev-blocker-rpc-to-advance-state-of-the-art-backrunning-auction-infrastructure) in January 2026) |
| **Contract** | [gnosis/cow-token](https://github.com/gnosis/cow-token) (COW / vCOW) |

**CoW DAO** is the community that governs [CoW Protocol](https://cow.fi/), a trading protocol built around _intents_ rather than direct swaps. It is one of the cleanest examples of [GnosisDAO](/wiki/daos/networks/gnosisdao)'s "on-chain holding company" pattern: CoW (originally "CowSwap") began as an internal Gnosis experiment, then graduated into a standalone project with its own token and treasury via [GIP-13 (Phase 2)](https://forum.gnosis.io/t/gip-13-phase-2-cowdao-and-cow-token/2735) and a [$23M round](https://www.theblock.co/linked/139692/cow-protocol-raises-23-million-and-spins-out-from-gnosis-dao) in January 2022. The name is a pun on _Coincidence of Wants_ — when two traders want opposite sides of the same trade, they can be matched directly against each other instead of both paying an AMM.

## Intents and batch auctions

Most DEXs execute one swap at a time against a liquidity pool, which exposes every trade to [MEV](https://ethereum.org/developers/docs/mev/) — front-running, sandwiching, and priority-gas auctions that quietly tax the trader. CoW Protocol restructures this. A user signs an _intent_ ("I want at least X of token B for my token A") rather than a transaction, and intents are collected into discrete **batch auctions**. Independent third parties called _solvers_ then compete to settle each batch: they net off any [coincidences of wants](https://docs.cow.fi/) peer-to-peer, route the remainder across on-chain liquidity, and the winning solver is the one that returns the most surplus to users. Every trade in a batch clears at the same [uniform price](https://docs.cow.fi/cow-protocol/concepts/introduction/fair-combinatorial-auction), which removes the ordering advantage MEV bots rely on. The protocol also ships [CoW AMM](https://cow.fi/cow-amm), a liquidity-provision design that protects LPs from loss-versus-rebalancing. It also co-created [MEV Blocker](https://mevblocker.io/), an RPC endpoint that shields ordinary wallets from sandwich attacks, but no longer runs it: the RPC was [acquired](https://cow.fi/learn/special-mechanisms-group-acquires-mev-blocker-rpc-to-advance-state-of-the-art-backrunning-auction-infrastructure) on 23 January 2026 by Special Mechanisms Group, the blockchain-microstructure division of Consensys, with the DAO supporting the transition and keeping [MEV protection](/wiki/economics/mev-and-value-extraction) in CoW Protocol and CoW Swap. The same intent-and-solver shape appears in the [1inch DAO](/wiki/daos/dexs/1inch-dao)'s Fusion mode, where competing resolvers settle gasless orders through Dutch auctions.

## Governance: COW, CIPs, and Snapshot

CoW DAO controls the protocol treasury — **44.4%** of the [one-billion fixed COW supply](https://docs.cow.fi/governance/token) — plus solver approvals, fee parameters, and grants. Governance follows the same optimistic pattern its parent [GnosisDAO](/wiki/daos/networks/gnosisdao) pioneered: ideas are debated as **CoW Improvement Proposals (CIPs)** on the [CoWmunity forum](https://forum.cow.fi/), then ratified by a gas-free [Snapshot](https://snapshot.org/#/cow.eth) vote — [off-chain signalling](/wiki/dao-governance/concepts/fundamentals/on-chain-vs-off-chain-governance) that costs holders nothing. Team and investor allocations were issued as non-transferable **vCOW** that [vests linearly over four years](https://docs.cow.fi/governance/token) and converts 1:1 to COW, aligning insiders to the long term. Because 10% of supply sits with GnosisDAO, the two DAOs remain economically intertwined — a recurring feature of the [spun-out-subsidiary](/wiki/dao-governance/concepts/treasury/dao-treasury-management) model.

## The cow.fi domain hijack and CIP-86 (2026)

On **14 April 2026** CoW Protocol's domain registrar (Gandi) was [compromised via social engineering](https://ambcrypto.com/cow-dao-approves-voluntary-refunds-despite-no-protocol-breach-in-domain-hijack/), letting attackers hijack DNS for the `cow.fi` domain for roughly 4.5 hours and serve a phishing UI that drained wallets of an estimated ~$1.2M. The protocol's smart contracts were never breached — the failure was entirely at the Web2 edge. What happened next is the governance story worth studying: rather than disclaim responsibility for an off-chain attack, the DAO passed [CIP-86](https://forum.cow.fi/t/cip-86-discretionary-grants-program-for-victims-of-the-cow-fi-domain-hijacking/3431), a [discretionary grants program](https://crypto.news/cow-dao-approves-compensation-for-cow-fi-hijack-victims-claims-due-may-14/) reimbursing eligible phished users up to 100% of losses from the treasury. The DAO was explicit that the payments were voluntary "goodwill" grants, not an admission of legal liability — a deliberate framing that let a treasury-funded reimbursement pass [on Snapshot](https://www.cryptotimes.io/2026/05/12/cip-86-passed-cow-dao-begins-compensation-for-april-attack/) without setting a liability precedent. It is a clean example of a DAO treating its treasury as a stakeholder-protection backstop.

## How Caper approaches this

CoW DAO's CIP-86 vote is a reminder of what a treasury is _for_ — a shared pool that can be deployed to protect the people who make the protocol valuable. But reaching it required a bespoke proposal, a Snapshot quorum, and a discretionary grants process stood up after the fact. [Caper](/wiki/foundations/what-is-a-caper) bakes the members-first claim on the treasury into the protocol itself: every caper's `exit()` lets a member who has voted or traded redeem a pro-rata slice of the treasury at any time, with no proposal or quorum, and that redemption share equals their exact [canonical vote weight](/wiki/governance/voting) `(t·v)/(V·T)` — the influence you hold and the value you can walk away with are the same number. Where CoW DAO had to _vote_ its treasury into acting for members, Caper makes the members' standing claim a protocol default. (Caper does not attempt CoW's intent-based trade execution — the parallel is purely about treasury governance.)