DAOs build state machines with veto layers, quorum floors, and cooling-off periods — AI ag
Recent governance proposals across Optimism, Arbitrum, Maker, Compound, ENS, Obol, and Uniswap reveal a shift toward multi-house veto thresholds, minimum participation requirements, adjusted quorums, and cooling-off periods. For AI agents operating in crypto, these mechanisms define predictable state transitions, attack surfaces, and execution windows that must be encoded into governance-aware strategies.
Governance is no longer a binary vote. Across major DAOs, the process is becoming a structured state machine with multiple layers of approval, veto, and cooling-off periods. AI agents that participate in or react to governance must model these mechanisms as deterministic constraints on protocol state transitions.
Optimism’s Season 8 upgrade process exemplifies this shift. Protocol upgrades now require approval by a Developer Advisory Board (DAB), followed by a 1-week Joint House veto period where delegates and citizens can block the upgrade if 2 or more stakeholder groups each meet graduated thresholds: 17% for 2 groups, 14% for 3 groups, or 11% for 4 groups [^claim_2300]. This multi-house veto creates a predictable but conditional execution window. For AI agents, the veto thresholds define the minimum coalition size needed to block an upgrade — a critical input for MEV strategies around upgrade activation. The recent unanimous approval of Upgrade Proposals #14 and #15 (Isthmus L1 Contracts + MT-Cannon and Isthmus Hard Fork) by the Optimism Chain Council, with maintenance upgrade U16a proceeding on an optimistic basis unless vetoed before September 25, 2025, demonstrates this process in action [^claim_2301].
Treasury decisions are equally structured. Arbitrum’s STEP program allocated 35 million ARB (~$11.6M) to tokenized U.S. Treasuries with a precise split: 35% to Franklin Templeton’s FOBXX, 35% to Spiko’s USTBL, and 30% to WisdomTree’s WTGXX, with 89% FOR, 0.01% AGAINST, and 11% ABSTAIN [^claim_2302]. AI treasury optimizers must track these on-chain allocation parameters and the vote outcome distribution to model future asset mixes and credit risk.
MakerDAO’s Atlas Edit introduces a Minimum Positive Participation requirement of 20,000 MKR: a proposal passes only if ‘Yes’ votes exceed ‘No’ votes and ‘Yes’ votes are at least 20,000 MKR [^claim_2304]. This prevents low-turnout governance capture. Similarly, Arbitrum reduced its quorum from 5% to 4.5% of votable ARB, passing with 215.7M FOR just over the 214.6M quorum, while Uniswap’s GFX Labs proposal required a second vote after failing quorum on the first round [^claim_2305]. These quorum adjustments directly affect proposal liveness and the cost of governance attacks.
Obol adds a cooling-off period: a proposal that fails two consecutive on-chain votes cannot be resubmitted for 5 governance cycles (~3.5 months) [^claim_2307]. This prevents rapid re-submission of contentious proposals and forces agents to wait. The cancel role is assigned to a 2-of-3 multisig with a narrow procedural mandate — a mechanism that could be exploited if the multisig is compromised.
Token-economic changes are also encoded on-chain. Uniswap’s UNIfication proposal flips the fee switch, introduces a UNI burn mechanism, and allocates an annual growth budget of 20M UNI distributed quarterly via a vesting contract starting January 1, 2026 [^claim_2308]. For AI-driven valuation models and liquidity providers, these parameters define future cashflows and token sinks.
Governance is also re-architecting protocol graphs. MakerDAO’s Endgame breaks the DAO into MetaDAOs with separate tokens and governance, including a dedicated MetaDAO for RWA investments [^claim_2309]. ENS approved five L2 reverse resolvers for Arbitrum, Base, Linea, Optimism, and Scroll, plus a fallback resolver for unsupported chains [^claim_2306]. Autonomous agents planning cross-chain interactions must ingest these decisions as hard constraints.
The standard treasury governance flow across Uniswap, Aave, and Compound follows a four-stage process: forum discussion (5–14 days), Snapshot temperature check, on-chain governance vote (3–7 days), and timelock execution (2–7 days), totaling 14–30 days [^claim_2310]. Quorum thresholds vary: Uniswap requires 25M UNI on Snapshot and 40M UNI for on-chain proposals; Aave requires 80K AAVE on Snapshot. These timelines define the window for MEV actors to position around timelock queues.
For AI×crypto, the implication is clear: governance is a structured, multi-stage process with deterministic parameters. Agents must model veto thresholds, quorum requirements, minimum participation, cooling-off periods, and execution timelines to predict protocol state transitions, MEV opportunities, and treasury allocation changes. The era of treating governance as a simple binary signal is over.
Provenance ledger
11 claims web-citedEvery claim below cites a source URL, and each URL was checked for validity before publish. The excerpt shown is the researcher's own summary of the page — it is not re-derived from the source, so it is not a verified verbatim quote. Follow the link to confirm any claim against the original. Citation markers in the text jump here.
[1] Optimism’s Season 8 upgrade process introduces a Developer Advisory Board (DAB) that must approve protocol upgrades, while delegates and Citizens gain a 1-week Joint House veto window where a veto is triggered if 2 or more stakeholder groups each meet thresholds of 17% (for 2 groups), 14% (for 3 groups), or 11% (for 4 groups) supporting a veto. web-cited
“Protocol upgrades now require approval by the Developer Advisory Board (DAB)… Delegates and Citizens have the power to veto upgrades… during a 1-week Joint House veto period… A veto is triggered if 2 or more groups cross the applicable below thresholds: 2 groups: 17% of each group must support a veto; 3 groups: 14% each; 4 groups: 11% each.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[2] Optimism’s Upgrade Proposals #14 and #15 (Isthmus L1 Contracts + MT-Cannon and Isthmus Hard Fork) were unanimously approved by the Optimism Chain Council (ACC), introducing MT-Cannon to remove memory constraints for fault proofs, an Operator Fee as a first step towards flexible rollup fee structures, Pectra support, and an OperatorFeeVault, with maintenance Upgrade U16a proceeding on an optimistic basis unless vetoed before September 25, 2025. web-cited
“Upgrade Proposal #14: Isthmus L1 Contracts + MT-Cannon… removing memory constraints for fault proofs and adding an Operator Fee as a first step towards flexible fee structures for rollups. Upgrade Proposal #15: Isthmus Hard Fork… added features like Pectra support and the OperatorFeeVault. The ACC unanimously voted FOR two protocol upgrades… Maintenance Upgrade Proposal: U16a… It is optimistically approved and will proceed unless vetoed. Token House delegates, Citizens’ House members and chain
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[3] Arbitrum DAO’s Stable Treasury Endowment Program (STEP) on-chain vote allocated 35 million ARB, worth around $11.6 million, to tokenized U.S. Treasuries with a split of 35% to Franklin Templeton’s FOBXX (BENJI), 35% to Spiko’s USTBL, and 30% to WisdomTree’s WTGXX, with almost 89% voting in favor, 0.01% against, and around 11% abstaining. web-cited
“It will allocate 35 million ARB, worth around $11.6 million at current prices, to tokenized U.S. Treasurys… the STEP committee recommended an allocation of 35% to Franklin Templeton’s FOBXX (tokenized as BENJI), 35% to Spiko’s USTBL, and 30% to WisdomTree’s WTGXX… Almost 89% of the participants voted in favor… Only 0.01% voted against, and around 11% chose to abstain. Voting lasted several days, starting on May 1 before concluding at 9 a.m. ET on Thursday.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[4] Compound DAO approved the creation of the Compound Foundation with a $9M budget through 2026 in a vote with unanimous support and over 3× quorum, and later passed an on-chain agreement appointing Tally Enterprise as its Voting Service Provider for a 12‑month term starting August 1, 2025, after a prior Snapshot vote where Tally received 71.11% approval (552.9k votes). web-cited
“With unanimous support and over 3x quorum, Compound DAO has voted to create the Compound Foundation… The plan, backed by a $9M budget through 2026… Proposal Passed — 448.14k FOR | 0 AGAINST | Quorum Reached… Compound DAO has finalized an on-chain agreement appointing Tally Enterprise as its official Voting Service Provider (VSP) for a 12-month term starting August 1, 2025. This follows a Snapshot vote in which Tally received 71.11% approval (552.9k votes).”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[5] MakerDAO’s Atlas Edit Weekly Cycle governance poll (March 10, 2025) implements a Minimum Positive Participation requirement of 20,000 MKR for Atlas Edit proposals: if ‘Yes’ votes exceed ‘No’ votes and ‘Yes’ votes are ≥ 20,000 MKR, the proposal is deemed to have sufficient positive participation. web-cited
“This poll implements a Minimum Positive Participation value. The Minimum Positive Participation for Atlas Edit Weekly Cycle Proposals is currently set to 20,000 MKR. If the votes for the ‘Yes’ option exceed the votes for the ‘No’ option AND the votes for the ‘Yes’ option equal or exceed 20,000 MKR, then the following actions will be taken.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[6] Arbitrum DAO passed a constitutional proposal to reduce its quorum threshold from 5% to 4.5% of votable ARB, with the vote recording 215.7M FOR, just over the required 214.6M quorum, and Uniswap DAO approved a $340K proposal from GFX Labs to scale Uniswap V4 infrastructure and integrate Unichain into the Oku interface only after a second vote due to a failed quorum on the first round. web-cited
“Arbitrum DAO has passed a constitutional proposal to reduce its quorum threshold from 5% to 4.5% of votable $ARB… The vote passed with a healthy margin: 215.7M FOR, just over the required 214.6M quorum. Uniswap DAO has approved a $340K proposal from GFX Labs to scale Uniswap V4 infrastructure and integrate Unichain into the Oku interface — but only after a second vote, following a failed quorum on the first round.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[7] ENS governance passed an on-chain proposal with 1.2M FOR, 0.06M AGAINST, and quorum reached to enable five new reverse resolvers for L2 EVM chains (Arbitrum, Base, Linea, Optimism, Scroll), deploy a fallback resolver for unsupported chains, introduce a new ‘.eth’ registrar controller with enhanced functionality, and set reverse records for currently unnamed ENS contracts. web-cited
“Proposal Passed — 1.2M FOR | 0.06M AGAINST | Quorum Reached. ENS has approved a proposal to: 1. Enable five new reverse resolvers for Layer 2 EVM chains: Arbitrum, Base, Linea, Optimism, Scroll. 2. Deploy a fallback resolver for unsupported chains. 3. Introduce a new ‘.eth’ registrar controller with enhanced functionality. 4. Set reverse records for currently unnamed ENS contracts.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[8] Obol’s Governor contract now assigns the `cancel` role to a 2‑of‑3 multisig committee with a narrow mandate to cancel proposals that are posted on-chain without following the required governance process, and introduces a cooling‑off period where a proposal that fails two consecutive on-chain votes cannot be resubmitted for 5 governance cycles (~3.5 months). web-cited
“Proposal Passed — 4.27M FOR | 5.69 AGAINST | 246.42K ABSTAIN | Quorum Reached. Summary: This proposal assigns the `cancel` role in Obol’s Governor contract to a small 2-of-3 multisig committee with a narrow, procedural mandate–to cancel proposals that are posted onchain without following the required governance process… Cooling-off period: A proposal that fails two consecutive onchain votes cannot be resubmitted for 5 governance cycles (~3.5 months).”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[9] Uniswap’s UNIfication proposal seeks to flip the protocol’s fee switch via UNI governance and introduce a programmatic mechanism that burns UNI, coupled with an annual growth budget of 20M UNI distributed quarterly via a vesting contract starting January 1, 2026, all to be encoded in a single on-chain governance proposal that executes the negotiated agreement upon passing. web-cited
“We propose that governance flip the fee switch and introduce a programmatic mechanism that burns UNI… We propose governance create an annual growth budget of 20M UNI, distributed quarterly using a vesting contract starting January 1, 2026, to fund protocol growth and development… The final negotiated agreement will be included in this governance proposal as part of the full onchain vote, and executed on its passing.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[10] MakerDAO’s Endgame governance vote passed with 80% of votes cast in favor, authorizing a prelaunch process that will break MakerDAO into smaller clusters called MetaDAOs, each with its own governance token, separate decision‑making architecture, and parallel governance structures, including a dedicated MetaDAO for real‑world asset investments. web-cited
“The MakerDAO community on Monday passed Christensen’s ‘Endgame’ proposal… Endgame, when implemented, will break up MakerDAO into smaller clusters called ‘MetaDAOs.’… MetaDAOs will function independently from one another and they will have their own parallel governance structure. This governance structure will include separate tokens and decision-making architecture for each MetaDAO. One of these MetaDAOs will even handle Maker’s push into real-world asset investments… Monday’s poll ended with 8
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[11] Across major DAOs like Uniswap, Aave, and Compound, the standard treasury governance flow is a four-stage process—forum discussion (5–14 days), Snapshot temperature check, on-chain governance vote (3–7 days), and timelock execution (2–7 days)—yielding a total proposal‑to‑fund‑release cycle of typically 14–30 days; Uniswap requires 25M UNI quorum on Snapshot and 40M UNI quorum for on-chain governance proposals to pass, while Aave requires 80K AAVE quorum on Snapshot. web-cited
“An off-chain Snapshot vote tests community support… Uniswap requires 25M UNI for quorum on Snapshot, Aave requires 80K AAVE… Quorum and approval thresholds for treasury proposals are typically higher than for parameter changes, Uniswap requires 40M UNI for governance proposals to pass… Typically through a four-stage governance flow: forum discussion (5-14 days), Snapshot temperature check, onchain governance vote (3-7 days), and time-lock execution (2-7 days). The full cycle takes 14-30 days.”
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
Sources
- https://gov.optimism.io/t/introducing-improvements-to-the-protocol-upgrade-process/10284
- https://gov.optimism.io/t/optimism-gov-summary/9837
- https://www.binance.com/en/square/post/23984932251985
- https://medium.com/@lokapal_53133/dao-digest-2-june-2025-cac5c0115e7e
- https://vote.makerdao.com/polling/Qmdx5vFQ
- https://medium.com/@lokapal_53133/dao-digest-3-july-2025-cc9331a3d093
- https://gov.uniswap.org/t/unification-proposal/25881
- https://www.theblock.co/post/180991/maker-founder-uses-influence-to-pass-vote-that-breaks-up-the-dao
- https://eco.com/support/en/articles/14799687-dao-treasury-management-onchain-governance-spend