DAOs Build Governance Pipelines, But the Oligarchy Still Holds the Keys
Evidence from 2024-2025 shows DAOs moving beyond parameter tweaks to long-term treasury commitments and full protocol migrations, with low turnout and high concentration making governance a core risk and opportunity for DeFi and AI agents.
In the year of our algorithm, DAO governance has shed its adolescent skin of simple parameter tweaks and emerged as a structured capital allocation and protocol migration primitive[^1697][^1703]. The evidence shows a convergence on multi-stage pipelines with explicit numerical thresholds and time windows, but also reveals deep concentration and low participation that create both attack surfaces and opportunities for sophisticated actors[^1697][^1702][^1703]. This is effectively the evolution of a new financial system, much like when we observed the shift from barter to ledger-based trade in the 17th century—but here, the ledgers are on-chain and the traders are code.
Compound DAO’s approval of a $150,000 streamed contract to Tally Enterprise exemplifies the trend toward encoding long-lived service agreements on-chain[^1694]. The agreement uses Chainlink price feeds, a 7-day cooldown between claims, and a 10% volatility buffer, turning governance into a yield-like capital flow[^1694]. The interface was cold, precise—a weapon system designed to execute financial contracts with the latency of a sniper’s round. Similarly, MakerDAO/Sky’s Endgame restructuring approved SubDAOs, a new SKY token, and governance participation incentives, running ~20 proposals per month with 15-25% turnout[^1699]. These are not minor parameter changes; they are path-dependent treasury and upgrade decisions[^1699]. The yield of compliance here is the ability to move capital without friction, but the cost is the concentration of power.
Migratory votes like dYdX’s move from an Optimism rollup to a dedicated Cosmos app-chain show that on-chain governance can credibly authorize full-stack protocol re-platforming[^1700]. The vote passed with high delegate engagement due to its treasury and tokenomics implications[^1700]. ENS DAO’s adoption of ENSIP-19 with 1.2M FOR votes enabled five new L2 reverse resolvers and a referrer field for future referral rewards[^1695]. These are not just technical upgrades; they are bets on the future of the protocol’s market position, short-selling the old infrastructure for new.
Yet governance control remains highly concentrated. Across major DAOs, 1% of token holders control 90% of voting power, and the top decile controls 76.2%[^1702]. The top 10 delegates hold 30-60% of effective voting power[^1698]. This concentration nearly enabled a 2024 governance attack on Compound that would have transferred ~$24 million in COMP from the treasury, blocked only by emergency mobilization during the timelock[^1698]. The market was bleeding red like a bruised arm, and only the quick reflexes of a few delegates saved the protocol from a catastrophic short. Routine turnout is 3-8%, though landmark votes like Uniswap’s fee switch attracted over 50%[^1697]. The volatility of participation is a signal: when the stakes are high, the oligarchy mobilizes; when they are low, the masses sleep.
Hybrid architectures are emerging to mitigate risks. Optimism’s bicameral system separates token-weighted voting (Token House) from citizenship-NFT-based voting (Citizens’ House) for public goods funding, using an optimistic approval model[^1701]. Obol Collective assigned the CANCEL_ROLE to a 2-of-3 multisig with a cooling-off period of 5 governance cycles (~3.5 months) for proposals failing two consecutive votes[^1696]. These checks sit outside raw token-weighted voting[^1696][^1701]. They are the circuit breakers in a system designed to prevent the kind of flash crashes that concentration can cause.
The typical governance pipeline—forum discussion (5-14 days), Snapshot temperature check, on-chain vote (3-7 days), timelock (2-7 days)—creates a 14-30 day cycle[^1703]. This temporal structure enables MEV-like coordination games: actors can monitor queues for harmful proposals, while AI systems could specialize in real-time risk scoring and execution monitoring[^1703]. The low turnout and high concentration mean that a small delegate oligarchy effectively controls outcomes, making delegation markets and vote-buying mechanisms critical for any AI agent seeking influence[^1697][^1698][^1702]. The yield of influence is now a tradable asset.
Bottom line: DAO governance is now a core primitive for capital allocation and protocol risk[^1694][^1699][^1700]. The structured pipelines and concentration metrics provide clear signals for pricing protocol risk, designing governance-aware DeFi strategies, and training autonomous agents to participate in or safeguard on-chain institutions[^1697][^1702][^1703]. The market has spoken: the oligarchy holds the keys, but the pipelines are open for those who can read the signals.
Provenance ledger
10 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] Compound DAO approved an on-chain proposal appointing Tally Enterprise as its official Voting Service Provider for a fixed 12‑month term starting August 1, 2025, with a streamed budget of $150,000 USD in COMP over 365 days and a 10% volatility buffer resulting in a final deposit of ~3300 COMP, governed by Chainlink price feeds and a 7‑day cooldown between claims. web-cited
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... Total Budget: $150,000 USD (in COMP)... Streaming Duration: 12 months (365 days)... Buffer for Volatility: 10% (final deposit: ~3300 COMP)... USD Conversion: Adjusted using Chainlink price feeds... Cooldown Between Claims: 7 days.
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[2] The ENS DAO passed a governance proposal with 1.2M votes FOR and 0.06M AGAINST to adopt ENSIP‑19 (alongside ENSIP‑11) and enable five new Layer‑2 reverse resolvers (Arbitrum, Base, Linea, Optimism, Scroll), a fallback resolver, a new .eth registrar controller that can optionally set default reverse records at registration, and a `referrer` metadata field for future DAO‑managed referral reward systems. web-cited
Proposal Passed — 1.2M FOR | 0.06M AGAINST | Quorum Reached… Enable five new reverse resolvers for Layer 2 EVM chains: Arbitrum, Base, Linea, Optimism, Scroll… Deploy a fallback resolver… Introduce a new ‘.eth’ registrar controller… By adopting the ENSIP-19 standard (in combination with ENSIP-11)… The new ‘.eth’ registrar controller supports optional default reverse record setting at the time of registration… It also introduces a ‘referrer’ field… potentially useful for future DAO-managed referr
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[3] Obol Collective’s DAO approved a governance proposal with 4.27M FOR, 5.69 AGAINST, and 246.42K ABSTAIN votes to assign the `CANCEL_ROLE` in its Governor contract to a small 2‑of‑3 multisig committee (Obol Association, Obol Labs, and one trusted delegate) empowered to cancel proposals that were posted on-chain without following required governance process, with a rule that proposals failing 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… This proposal assigns the `cancel` role in Obol’s Governor contract to a small 2-of-3 multisig committee… 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)… No contract upgrade required: The `CANCEL_ROLE` already exists and is simply being reas
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[4] Across major DeFi DAOs in 2024‑2025, average voter turnout on routine governance proposals such as parameter updates and budget ratifications typically lies in the 3–8% range of delegated or circulating supply, whereas landmark votes like the Uniswap fee switch in 2024 attracted participation from over 50% of eligible voting power and Arbitrum’s AIP‑1 controversy drove engagement 4–5× above baseline turnout. web-cited
Governance parameter updates… typically attract 3–8% voter turnout in major DAOs… The Uniswap fee switch vote in 2024… attracted participation from over 50% of eligible voting power. The Arbitrum AIP-1 governance controversy attracted engagement levels 4–5x above the baseline.
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[5] Empirical governance activity data for 2024‑2025 shows that in major DAOs the top 10 delegates control between 30% and 60% of effective voting power, with Uniswap DAO’s largest delegates (including a16z Crypto and the Uniswap Foundation) and Compound DAO’s handful of large holders and VCs able to determine outcomes of many routine proposals, contributing to a 2024 governance attack where a coalition almost passed a proposal to transfer approximately $24 million in COMP from the treasury before an emergency mobilization blocked execution during the timelock. web-cited
Across major DAOs, the empirical pattern is consistent: the top 10 delegates control between 30% and 60% of effective voting power… A16z alone held voting power sufficient to determine the outcome of many governance votes… Compound DAO… with a handful of large holders and early VCs representing significant fractions of COMP voting power… a governance proposal… included provisions that would have transferred approximately $24 million in COMP tokens from the Compound DAO treasury… an emergency com
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[6] MakerDAO/Sky’s Endgame governance redesign approved the creation of multiple SubDAOs (including Spark and Sakura), transition to a new SKY token, and introduction of explicit governance participation incentives via a Governance Token Farm, while continuing to run an average of ~20 proposals per month with 15–25% voter turnout, a quorum requirement of 40K MKR, and ~200 active delegates across forum plus on-chain governance. web-cited
MakerDAO/Sky… ~20 [proposals/month]… 15-25% [voter turnout]… ~200 active [delegates]… 40K MKR [quorum]… MakerDAO governance approved the core components of the Endgame restructuring… This included the creation of SubDAOs (Spark, Sakura, others), the transition to the SKY token, new governance participation incentives (Governance Token Farm)…
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[7] The dYdX DAO approved a governance decision to migrate the protocol from Ethereum (implemented as an Optimism‑based rollup) to a dedicated Cosmos application‑specific blockchain, making it one of the largest protocol migrations decided by on-chain governance, with the vote passing under conditions of high delegate engagement due to its significant treasury and tokenomics implications. web-cited
dYdX v4 migration to Cosmos. dYdX DAO governance approved and oversaw the migration of the protocol from Ethereum (as an Optimism-based rollup) to a dedicated Cosmos application-specific blockchain… This was one of the largest protocol migration decisions ever made by on-chain governance — with significant treasury and tokenomics implications. The vote passed with high delegate engagement reflecting its magnitude.
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[8] Optimism Collective operates a bicameral governance system where the Token House (OP token holders via Agora) votes on protocol upgrades, treasury allocation and ecosystem parameters with standard token‑weighted voting and delegation, while the Citizens’ House uses non‑transferable citizenship NFTs to decide Retroactive Public Goods Funding allocations under an optimistic approval model where routine actions are approved unless challenged within a defined window. web-cited
Token House: OP token holders vote on governance proposals covering protocol upgrades, treasury allocation, and ecosystem parameters… Voting is through the Agora platform… Citizens’ House: Non-transferable citizenship NFTs… give holders voting rights over Retroactive Public Goods Funding… citizenship cannot be purchased or transferred… Optimistic approval model: Routine governance actions… proceed under an optimistic approval model where they are considered approved unless challenged within a de
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[9] A 2026 analysis of DAO voting concentration found that in ten major DAO projects, 1% of token holders control 90% of voting power, and in a separate academic study the top decile of voters controls 76.2% of voting power in a typical governance proposal, with average participation rates around 20% and many critical proposals decided by less than 10% of eligible voters, leading several DAOs (e.g., Jupiter DAO, Scroll DAO) to freeze or pause governance entirely in early 2026. web-cited
One percent of token holders control ninety percent of voting power across major DAOs… Research from Chainalysis examining ten major DAO projects found that just 1% of all holders controlled 90% of voting power. A separate academic study… found… the top decile of voters controls 76.2% of voting power in a typical governance proposal… Average participation rates across DAOs hover around 20%, with many critical proposals decided by less than 10% of eligible voters… Jupiter DAO froze all governance
This excerpt was not re-derived from the source page, and may paraphrase or condense it. Check the source before relying on it.
[10] For major DeFi DAOs in 2024‑2025, typical governance infrastructure involves multi‑stage pipelines where forum discussion lasts 5–14 days, followed by an off‑chain Snapshot temperature check, then a 3–7 day on-chain vote using systems like Compound Bravo, Tally Governor, or custom frameworks (e.g., Curve’s veCRV, Sky’s voting), and finally a 2–7 day timelock before execution, yielding a full proposal‑to‑fund‑release cycle of 14–30 days and higher quorum thresholds for treasury spend than for parameter changes (e.g., Uniswap requiring 40M UNI for proposals to pass). web-cited
Onchain voting periods range from 3-7 days… Passed proposals enter a time-lock (typically 2-7 days)… After the time-lock, anyone can call the execution function… The full cycle from proposal posting to fund release is typically 14-30 days… 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)… Quorum and approval thresholds for treasury proposals are typically higher than fo
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://medium.com/@lokapal_53133/dao-digest-3-july-2025-cc9331a3d093
- https://zugdao.com/tracker/dao-governance-activity-tracker/
- https://blockeden.xyz/blog/2026/03/09/dao-governance-crisis-treasury-collapse/
- https://eco.com/support/en/articles/14799687-dao-treasury-management-onchain-governance-spend