Aave's internal strife escalates: Behind the $51 million funding dispute lies a battle over "who defines history."

Written by: Castle Labs

Compiled by: Chopper, Foresight News

Aave is the largest lending protocol in the DeFi space, with a total value locked of $26 billion and annual revenue of $140 million, accounting for 60% of the DeFi lending market. Today, at its governance forum, a heated debate erupted regarding how it arrived at its current state.

The founding company Aave Labs and the most powerful representatives of the decentralized autonomous organization Aave DAO have released conflicting versions of the same historical events. Neither side is neutral; both wield significant governance power and have vested interests in the current token holder voting results.

The following is a factual account of the reconstruction based on the original file.

A proposal

On February 12, Aave Labs released a framework proposal titled "Aave Will Win." The proposal stipulates that 100% of product revenue will flow to the DAO, but Labs demands $42.5 million in stablecoins and 75,000 AAVE tokens in the first year, totaling approximately $51 million. This equates to 31.5% of the entire treasury and 42% of non-AAVE reserves.

The framework also proposes approving V4 as the technical future of the protocol, suspending new feature development for V3, and planning to eventually deprecate V3. Currently, V4 is still on the testnet, while all of Aave's revenue comes from V3.

Prior to the Snapshot vote, community feedback called for the public disclosure of wallet information, the establishment of a foundation structure before funding was distributed, and the linking of the deprecation of version V3 to the adoption milestone of version V4. However, the vote did not make any actionable commitments to any of these.

Two opposing narratives

On February 25, two posts appeared on the Aave governance forum within a few hours of each other.

Aave Labs has released a contribution report. The report covers all protocol versions from V1 to V4, flash loans, eMode, security modules, GHO, frontend, and branding. They have written over 570,000 lines of code since 2017. Regarding revenue attribution, their position is that the architecture implementing these strategies was their original design, and attributing revenue to any single contributor would distort how layered protocol development works.

ACI founder Marc Zeller released a financial analysis report. He pointed out that Aave Labs has accumulated a total of $86 million in funding, including the 2017 ICO, venture capital, DAO payments, and fee revenue he claims was misappropriated from DAOs without governance voting. He traced 23% of the token supply to 52 wallets connected to the founding infrastructure. He calculated that for every $1 in revenue generated by Labs' institutional-grade RWA product Horizon, the DAO had to pay approximately $24 in costs, with Merkl incentive spending at $4.2 million, while revenue was only $216,000. He also listed six independent products that he believes have failed or are still unprofitable, claiming that 98% of the revenue from version V3 did not come from code directly provided by Labs, but from code provided by BGD Labs and other DAO service providers.

Both sides have their own version of events, and both sides have their own interests at stake.

Core developers leave

Eight days after the release of the Aave Will Win framework, BGD Labs announced that it would not renew the contract after it expires on April 1.

BGD built versions V3.1 through V3.7, Liquid eMode, and most of Aave's governance infrastructure. Their reason for leaving is that Labs pressured V3 to roll out V4 without collaborating with BGD on V4 development and imposed what BGD calls "artificial restrictions" on improvements to V3. This team, which contributed all of Aave's current revenue-generating code, believes the current environment is no longer suitable for them. They proposed providing two months of security assurance to handle major security incidents. After June, they will leave permanently.

Since the brand ownership dispute in December 2025, AAVE's price has fallen by approximately 32%. During the same period, Morpho, which doesn't charge protocol fees, saw its price rise by approximately 42%. Whether this divergence is due to governance uncertainty or other factors remains unclear. However, one thing is certain: Aave generates $140 million in revenue annually and is conducting token buybacks, yet its token has underperformed its direct competitors by a significant margin over the past two months.

What exactly is the core of the controversy?

Setting aside the conflicting data, the core dispute is actually quite simple: Aave Labs is the initiator of the agreement and therefore deserves continued funding commensurate with its contributions. However, a coalition of DAO representatives argues that Labs is merely one of many service providers and should be subject to the same accountability standards as other service providers.

The two positions are consistent in themselves. This tension is not a dysfunction, but rather an indication that something unusual is happening within Aave.

Most DeFi protocols don't have this problem. Most protocols have a founding team that acts as a DAO, responsible for decision-making and fundraising. In most cases, governance is merely a formality. Aave is different: it boasts a genuine ecosystem of numerous technically strong and financially independent service providers (BGD, ACI, Chaos Labs, TokenLogic) who can genuinely challenge the founding team's proposals. This is extremely rare. Building such a distributed contributor infrastructure is incredibly difficult, and most protocols fail to achieve it.

The premise for the existence of conflict is that "Aave," however defined, has indeed succeeded.

The current issue is whether the DAO understood its own position before voting. BGD's departure is the clearest risk signal. If the governance environment becomes one where optimal independent developers leave due to pressure from the Labs route, then Aave's distinctive distributed contributor model may begin to crumble.