Historical source published February 26, 2026, at 19:22 UTC+8. Written by Castle Labs; Chinese translation by Chopper, Foresight News.
Aave’s governance dispute centers on more than a funding request. It asks whether the protocol’s founding company should receive continuing support on the strength of its foundational role—or be evaluated under the same accountability standards as other DAO-funded contributors.
At the time of the source report, Aave was described as DeFi’s largest lending protocol, with $26 billion in total value locked, $140 million in annual revenue and a 60% share of the DeFi lending market. Those figures make the dispute consequential, but they do not settle competing claims about who built that position or who should guide its next phase.
What the funding framework proposed
On February 12, Aave Labs published its “Aave Will Win” framework. It proposed directing 100% of product revenue to the DAO while requesting $42.5 million in stablecoins and 75,000 AAVE tokens for Labs in the first year—approximately $51 million in total.
According to the report, that request represented 31.5% of the treasury and 42% of its non-AAVE reserves.
The framework also sought approval for V4 as the protocol’s technical future, a halt to new feature development for V3 and V3’s eventual retirement. At the time, V4 remained on testnet, while V3 generated all of Aave’s revenue.
Before the Snapshot vote, community feedback called for wallet disclosures, a foundation structure before funds were disbursed, and milestones linking V3’s retirement to V4 adoption. The report said the vote contained no enforceable commitments on those points.
Two interested parties, two accounts of Aave’s success
On February 25, Aave Labs and ACI founder Marc Zeller published sharply different accounts on the governance forum within hours of each other. Neither was a neutral observer: both held substantial governance influence and had interests tied to the vote’s outcome.
Aave Labs’ contribution report covered V1 through V4, flash loans, eMode, the safety module, GHO, front-end development and branding. Labs said it had written more than 570,000 lines of code since 2017. Its argument was that the original architecture enabled subsequent revenue-generating strategies, making attribution to any single contributor misleading in a layered protocol.
Zeller’s financial analysis challenged that account. He put cumulative funding received by Aave Labs at $86 million, including the 2017 ICO, venture funding, DAO payments and fee-sharing proceeds that he alleged had been diverted from the DAO without a governance vote. That last point was an allegation, not an independently established finding in the supplied report.
He also said he had traced 23% of the token supply to 52 wallets linked to founding infrastructure. On Horizon, Labs’ institutional real-world asset product, he calculated that the DAO spent roughly $24 for every $1 it earned, citing $4.2 million in Merkl incentives against $216,000 in revenue.
Those Horizon figures concern one product’s reported economics; they do not establish a conclusion about RWA adoption as a whole. More narrowly, they distinguish spending intended to attract activity from revenue actually captured by the DAO.
Zeller additionally listed six standalone products he considered failed or still loss-making. He claimed that 98% of V3 revenue came from code supplied by BGD Labs and other DAO service providers rather than code directly supplied by Aave Labs. Labs’ architectural argument and Zeller’s revenue attribution therefore offered competing ways to define contribution, not a settled accounting of credit.
BGD’s planned departure made the risk concrete
Eight days after the framework appeared, BGD Labs announced that it would not renew its contract when it expired on April 1.
BGD had built V3.1 through V3.7, Liquid eMode and much of Aave’s governance infrastructure. It said Labs was pushing V4 at V3’s expense, had not collaborated with BGD on V4 development and had imposed what BGD described as “artificial constraints” on V3 improvements.
BGD concluded that the environment was no longer suitable for its work. It offered a two-month safety backstop for major security incidents and said it would be fully gone after June.
The immediate governance risk was therefore not merely disagreement over historical credit. It was the prospective loss of an independent technical contributor responsible for major components of the revenue-producing protocol.
Token performance did not establish the cause
The report said AAVE had fallen approximately 32% since a brand-ownership dispute in December 2025. Over the same period, Morpho—which it described as charging no protocol fees—had risen approximately 42%.
The source explicitly left the cause unresolved: governance uncertainty might have contributed to the divergence, but other factors could also explain it. Aave’s reported $140 million in annual revenue and ongoing token buybacks had not prevented its token from substantially underperforming that direct competitor over the two-month period.
The underlying question: what does decentralization require?
The source’s central interpretation was that this conflict reflected something unusual about Aave: a contributor ecosystem capable of challenging its founder.
It identified BGD, ACI, Chaos Labs and TokenLogic as technically capable, financially independent service providers. In that reading, disagreement was not itself proof of governance failure. It was evidence that decision-making extended beyond the founding team.
But that model depends on contributors being willing to remain. If roadmap decisions and funding arrangements drive independent developers away, the distributed structure that allows meaningful scrutiny could weaken.
The funding debate thus turned on two linked questions: what compensation Aave Labs’ continuing contribution justified, and whether the DAO could support that contribution without eroding accountability or its broader developer base.
