Decentralized autonomous organizations collectively control over $26 billion in on-chain treasuries as of Q1 2026, up from $12 billion in 2024, according to aggregated governance data. Uniswap leads at $4.8 billion, followed by Sky/MakerDAO ($3.9 billion), Optimism ($2.1 billion), Arbitrum ($1.7 ...
"GnosisDAO raised $12.5 million in 2017 and now controls more than $200 million in assets without any further fundraise in between." — Lukas Schor, Co-founder, Gnosis
Decentralized autonomous organizations collectively control over $26 billion in on-chain treasuries as of Q1 2026, up from $12 billion in 2024, according to aggregated governance data. Uniswap leads at $4.8 billion, followed by Sky/MakerDAO ($3.9 billion), Optimism ($2.1 billion), Arbitrum ($1.7 billion), and Lido ($1.4 billion). The scale of assets under DAO governance now exceeds the AUM of many mid-tier hedge funds.
Yet the governance infrastructure managing these assets remains structurally weak. Voter participation on major platforms like Snapshot sits below 5% of eligible holders. For 17 of the 23 largest DAOs, the top 10 delegates hold enough voting power to pass proposals unilaterally. Vote-buying marketplaces have emerged that allow a single actor to acquire $6.5 million in governance weight for $10,000. Activist groups known as "RFV Raiders" systematically target DAOs whose tokens trade below treasury net asset value, forcing liquidation votes. The gap between the capital stored in DAO treasuries and the governance mechanisms protecting it represents a structural vulnerability the ecosystem has not resolved.
DAO treasuries have more than doubled since 2024. Over $35 billion in assets are now governed by DAOs, with approximately $26 billion held directly in on-chain treasuries, according to governance analytics data from Q1 2026. The composition is heavily concentrated: the top five DAOs — Uniswap, Sky/MakerDAO, Optimism, Arbitrum, and Lido — account for roughly $14 billion, or 54% of the total.
These treasuries hold a mix of native governance tokens, stablecoins, ETH, and increasingly, tokenized real-world assets. The concentration creates a specific governance risk: most treasury value sits in the DAO's own governance token, meaning any sell pressure from a governance vote to unlock treasury assets could be self-reinforcing.
The scale has attracted institutional-grade scrutiny. DAO treasury audit practices have become a dedicated discipline, with multisig verification, on-chain spending proposal tracking, and post-mortem accountability frameworks now standard at the largest protocols.
The governance systems controlling these billions operate with participation rates that would be considered illegitimate in any traditional governance context. On Snapshot, the most widely used off-chain voting platform, fewer than 5% of eligible token holders participate in votes. On-chain voting platforms like DAOStack and Aragon report sub-10% turnout.
Uniswap, despite its $4.8 billion treasury, typically sees 5-15% of UNI tokens participate in governance proposals. The protocol requires a quorum of 40 million UNI for passage, but quorum is met through concentrated delegate voting rather than broad participation.
The concentration compounds the problem. Across the 23 largest DAOs analyzed, the top 10 delegates in 17 of them controlled enough voting power to pass proposals without any additional participation. This effectively means a small group of power delegates — often venture capital firms, protocol insiders, and professional delegates — control governance outcomes while the vast majority of token holders remain passive.
According to governance analytics, getting 350 to 500 voters per proposal is considered a strong turnout in the current DAO environment. The apathy stems from multiple factors: information overload from frequent proposals, low direct incentive for participation, and the perception that outcomes are predetermined by large holders.
The Gnosis DAO treasury redemption battle illustrates the vulnerability created by the gap between treasury value and token market capitalization. In May 2026, activist investors known as "RFV Raiders" — named for their focus on realized fund value — proposed GIP-150, which would allow GNO token holders to redeem tokens for a pro-rata share of the DAO's $223 million treasury at approximately $170 per token, nearly 30% above GNO's market price of $131 at the time.
The proposal was submitted by a community member identified as Wismerhill, who argued that Gnosis Ltd., the operating company led by the project's original co-founders, had become a "cash sink" structurally misaligned with token holders. The core claim: Gnosis Ltd. received $30 million in DAO funding, but over two quarters in which the company disclosed revenue, total reported income was under $300,000.
Gnosis co-founder Lukas Schor countered that GnosisDAO raised $12.5 million in its 2017 ICO and now controls more than $200 million in assets without any additional fundraising, while building substantial value for the industry.
The vote, which ran until May 12, 2026, attracted 330,000 votes. As of reporting from The Defiant during the vote period, 65% of votes cast opposed the redemption proposal. The outcome swung after a whale intervened to counter the co-founder's voting bloc.
The RFV Raiders have a documented track record. In 2023, the same group targeted Rook (generating a reported 5x return), pushed through the Fei/Tribe wind-down, and forced Aragon to repurpose its treasury into a grants program. Their strategy exploits a fundamental arbitrage: when a DAO token trades below the per-token value of its treasury, activist voters can profit by forcing a redemption or liquidation event.
The Gnosis case prompted a defensive response from other protocols. Beefy Finance introduced a buyback system in April 2026 after its BIFI token fell below net asset value, explicitly to avoid attracting an RFV play.
Aave DAO's April 2026 vote to approve $25 million in funding for Aave Labs represents one of the largest single governance-directed grants in DeFi history. The proposal passed with approximately 75% support — 522,780 tokens in favor versus 175,310 against — after months of dispute over revenue sharing and governance control.
The $25 million is denominated in aEthLidoGHO stablecoins, with $5 million released immediately upon execution on April 13, 2026, and the remaining $20 million streamed over 6- and 12-month tranches. The package also includes 75,000 AAVE tokens, valued at approximately $7 million at the time, unlocked linearly over 48 months from the ecosystem reserve.
A notable structural feature: approximately $17.5 million of the total is earmarked for product incentives tied to measurable milestones for Aave Crypto App and Aave Pro. This milestone-contingent structure represents a departure from the blank-check grants that characterized early DAO spending.
The proposal was structured under the "Aave Will Win" framework, which separates individual funding and development decisions into discrete governance proposals rather than omnibus votes. According to CoinDesk reporting, the vote ended "months-long fight over who controls protocol revenue," making it as much a power struggle as a spending decision.
The 25% opposition is significant in context. DAO governance votes rarely see organized opposition exceed 10-15%. The dissent signals that a meaningful segment of AAVE holders view the Labs relationship as extractive rather than value-generating — a structural tension present in most DAOs where a for-profit entity develops the protocol while the DAO funds operations.
Lido DAO proposed spending up to 10,000 stETH (approximately $20 million) from its treasury to buy back LDO governance tokens in March 2026, according to The Block and CoinDesk reporting. The proposal arrived with LDO trading at a 70% discount to its two-year median LDO-to-ETH ratio, sitting at approximately 0.00016.
Lido framed the buyback on economic grounds rather than price support. Net protocol rewards had dropped approximately 20% over the relevant period, while costs improved 13% year-over-year and the protocol's effective take rate rose to 6.11% from 5%. The implication: the token's discount reflected market sentiment divergence from operating fundamentals.
Execution would proceed in 1,000 stETH batches, each requiring a separate Easy Track motion with a three-day objection period. The Lido Growth Committee was authorized to route trades through centralized exchanges including Binance, OKX, Bybit, Gate, and Bitget, each offering more than $100,000 in depth.
At then-current prices, the proposed buyback could absorb roughly 65 million tokens, approximately 8% of the circulating supply. The proposal was explicitly framed as a one-off, distinct from Lido's separate NEST automated buyback proposal introduced in November 2025.
The buyback serves a dual purpose: narrowing the discount to NAV and preempting the kind of RFV activist play that targeted Gnosis. A token trading at a persistent discount to its share of DAO treasury assets is an open invitation for value extraction strategies.
The emergence of vote-buying marketplaces adds a new dimension to DAO governance risk. LobbyFi, deployed across Arbitrum, Optimism, Scroll, and zkSync, operates as a marketplace where token holders can delegate voting rights to third parties in exchange for financial compensation. Voting power is acquired on a per-proposal basis through fixed-price purchases or auctions.
The platform's impact became visible in April 2025, when a user identified as hitmonlee.eth paid 5 ETH (approximately $10,000) through LobbyFi to acquire 19.3 million ARB tokens' worth of voting power — approximately $6.5 million in governance weight. The votes were used to elect CupOJoseph to Arbitrum's Oversight and Transparency Committee. The acquired voting power exceeded that of major delegates including L2Beat and Wintermute.
The economics are stark: $10,000 to influence governance decisions over a $1.7 billion treasury. The Arbitrum Foundation acknowledged the issue but stated it would not unilaterally ban vote-buying, instead calling for community-led discussion.
LobbyFi represents the formalization of a practice — vote delegation for compensation — that previously existed informally. The protocol pitches itself as enabling yield on idle governance tokens. The unintended consequence is a cost-efficient pathway for governance capture. According to DL News analysis, the rise of LobbyFi is part of a broader "GovernFi" trend that treats governance participation as a financial instrument rather than a civic function.
Some DAOs have begun experimenting with structural alternatives to pure token-weighted voting. Data from governance analytics firms indicates that DAOs transitioning from one-token-one-vote to delegated or quadratic models saw voter turnout increase from 2.8% to 11.4% on average, with proposal quality scores rising 34%.
Quadratic voting — where voting cost increases quadratically with the number of votes, limiting the influence of large holders — has been adopted by 15 major DAOs, including Gitcoin and Optimism. Adoption of quadratic voting jumped 30% between 2024 and 2025, according to industry data.
Additional models in production include conviction voting (1Hive), where votes decay over time unless reinforced, and veToken models, where governance weight requires locking tokens for extended periods. The latter creates economic cost for governance participation, theoretically filtering for long-term aligned participants.
According to Karma's delegation analytics, 42% of voters delegated at least some of their governance power in 2024, indicating a shift toward representative structures within ostensibly direct-democracy systems.
The trajectory suggests that the pure one-token-one-vote model, which dominates current DAO governance, is being gradually displaced. The likely default architecture, according to governance researchers, is a hybrid and layered system: a base layer of token voting with delegation and quorum management, overlaid with vote locking to weight long-term alignment.
The DAO governance model faces a fundamental mismatch: the assets under management have scaled to institutional levels while the governance mechanisms remain in a pre-institutional state. Sub-5% participation, concentrated delegate power, formalized vote-buying, and systematic activist exploitation of treasury-to-market-cap discounts are not theoretical risks — they are documented, recurring events.
The responses vary. Lido and Beefy pursue buybacks to close the NAV gap. Aave attempts milestone-contingent funding. Gnosis faces a direct shareholder-style activist campaign. Gitcoin and Optimism experiment with quadratic voting. None of these represent a systemic solution.
The economic value locked in DAO treasuries demands governance infrastructure of corresponding rigor. The current state — where $10,000 can buy $6.5 million in voting influence and 95% of token holders decline to participate — represents an implicit subsidy to whoever is willing to engage the governance system, whether aligned participants or extractive actors. The data shows the problem clearly. The solutions remain fragmented and early-stage.