Alexos, you’re right. It is disheartening to see the discussion devolve like this, and you’re also right that emotion is not a substitute for data. A debate of this importance should be grounded in facts.
However, the post from ACI is not a factual report and most of it looks copy and pasted from Claude or ChatGPT. It is a narrative, an imaginative story crafted to paint a specific picture. It selectively presents data, misrepresents things when convenient, and omits critical context to build its case. For a productive discussion to happen, we have to start from a place of truth.
Emilio’s post was not an attempt to refute every claim in that narrative. It was a response from someone who was personally named in the hit piece. If you want a report card of sorts, you can see a non-exhaustive but helpful list of Aave Labs’ contributions here.
I will not address every misrepresentation in the ACI post, as that would be an endless task. But to your point about facts, let’s look at a few clear examples that show a pattern of dishonesty and you can decide how much of the rest of the post should be taken as truth.
In late 2024, the space migrated to aavedao.eth under the Dolce Vita Extension. When aavedao.eth expired in December 2025, Labs re-acquired it for 0.7 ETH, regaining ultimate controller access. Labs now holds both aave.eth and aavedao.eth.
This is extremely misleading framing. It was ACI who let the aavedao.eth name expire, which introduced a critical security risk to the Snapshot. Anyone who owned the ENS would have admin control over Aave DAO’s Snapshot space. Fortunately we stepped in to secure it and are still willing to send it to wherever makes sense as long as all receiving parties are on board. The attempt to frame this as a hostile action, when ACI owned the name, let it expire, and then didn’t try (or didn’t know) to recoup it, is dishonest.
However, sneaky claims like this are not a one-off. On the CoW Swap thread, the same author claimed the new swap adapters were using Balancer/Morpho flashloans, therefore bringing volume to competitors. This was either a claim out of ignorance, which is concerning, or an intentional lie to spin a narrative, which is worse. It was super easy to check this onchain before posting.
If an author is willing to misrepresent relatively small and verifiable issues like these, what else are they willing to misrepresent?
Since launch on August 27, 2025, the DAO has spent ~$4.2M in direct Merkl incentives (peaking at $43K/day in November, currently $23K/day).
This, again, is an intentional misrepresentation. Ripple graciously worked with us to provide those incentives to help bootstrap the market. Framing partner contributions as a DAO cost is absolutely silly, particularly when it’s only applied to things Aave Labs works on. If anything, this eases the burden on the DAO and we should be extremely grateful for such things.
That said, there is an incentive budget approved by the DAO, which is primarily being handled by TokenLogic in order to boost GHO borrowing.
And while there is an intense focus on the revenue generation of Aave Horizon, I would say it’s less than a year old, exploring a brand new (but growing use case), and has successfully put Aave’s in the onchain institutional picture. Many institutions have reached out interested in the market and without it would likely have decided to focus on competitors. Part of building successful protocols requires positioning for the future and Aave Horizon does just that - with the help of other service providers within the DAO.
Would also add that by definition, tokenized securities like the ones used on Aave Horizon as collateral are naturally going to have fewer borrowers. Only those approved by the RWA issuer can actually own the RWAs and that list is going to be a tiny subset of the total DeFi userbase for now. But we expect it to get much larger in the next couple of years.
It’s also a clever rhetorical trick to pretend we said that $1 billion RWAs were on Aave Horizon. But that’s not what the post said, instead it said there are $1 billion RWAs on Aave. This is a total deposits number that includes all RWA’s as they’re defined by rwa.xyz (some of which are in other Aave markets). We’ve always counted “total deposits” this way and the author had no problem when that number was the $75 billion all-time high. Only now when it’s convenient is it a problem.
A frontend that breaks launches.
For anyone who builds things, this (again) is a disingenuous claim. Name any major company that has not had a non-critical error or visual issue on their website. ACI is not immune from such technical issues themselves. Just a couple days ago, they had an issue with the freeze-all-reserves function on zkSync, Metis, and Soneium. We didn’t publicly attack them for it, because things happen in development. Same thing happened with the PYUSD / RLUSD rewards, etc. Notably, aave.com has not suffered any major exploits or security issues, which is not something every frontend in our industry can say. That is more important than small issues that get fixed rather quickly.
That frontend has run for years, helping millions of users interact with Aave.
Trashing the Revenue Engine to Sell the Replacement
This whole section is just inflammatory for the sake of it. We haven’t trashed V3 to sell anything. We’ve actively promoted V3, its success, and helped push BD efforts behind the scenes to make sure V3 grows. Many of the comments in reference are just comparative in nature. When making comparisons, its only natural to frame things in context of what the reader will already be familiar with.
Regarding V4 materials, most readers will know and understand V3, so that’s what it makes the most sense to compare it to. Given I wrote most of the material, I always made an effort to pay homage to V3, its tech, and what it has accomplished. Not only that, but I’ve also written pieces both internally and for business development that helps sell V3 and its benefits too. V4 is awesome and should be promoted as such! But you don’t have to take my word for it:
While we did propose a timeline, it was just that, a proposal. This was not something we weren’t open to more opinions on which is why it has now been removed from the Temp Check. Our primary point was that at a certain threshold, maintaining both systems probably just won’t make sense. V3 can essentially exist inside V4 if that functionality is desired. We’re happy to continue that conversation in the coming months, see what happens when V4 is actually live, etc.
The same pattern: take credit for the foundation, dismiss the improvements, minimize the revenue impact.
I find this claim… interesting given ACI’s own transparency report credits all of Aave DAO’s revenue to themselves. Sure there’s a thank you or two in there, but in the end they just claimed that for every dollar of compensation they generated $28 in revenue. Doesn’t exactly seem like spreading the love. I also think it’s an extremely crude metric that doesn’t paint the full picture of how the DAO works together.
We also didn’t minimize Liquid eMode, like at all. We explicitly mention the importance of Liquid eMode and the importance of protocol evolutions in our contribution report:
”eMode was originally introduced as part of the V3 architecture, establishing the technical foundation for more capital-efficient strategies. In the v3.2 upgrade, BGD Labs expanded this framework through Liquid eMode, enabling a single asset to participate in multiple eMode categories simultaneously. Together, these architectural steps created the flexibility that underpins many of the protocol’s current revenue strategies. Those outcomes reflect cumulative contributions across protocol design, subsequent upgrades, risk analysis, asset onboarding, and DAO governance decisions.”
Our point was the opposite of what is implied in the piece. Which is that Aave is the result of a ton of group effort across the board.
Coinbase: Labs Failed to Compete
This section was pretty disappointing to read. Obviously, framing it like this makes it seem way worse. He knows there are professional, and sometimes legal reasons, stuff like this shouldn’t just be laid out in the open. Not sure it’s going to help anything in the future either. I think the professional thing to do is keep stuff like this behind close doors because lots go into negotiations. And sometimes, teams are already opinionated on the direction they want to go, an architecture they think suits their use case better, list goes on.
Hello, thanks for your comment, I want to point out that since labs escalated the AWW proposal to a vote without including any enforceable commitments from community feedback we’re acknowleging that the debate is pretty much over.
Again, it’s just so easy to verify this isn’t true. What is up for vote is a Temp Check. There are two more stages of voting after that before anything is confirmed. We are actively gathering feedback and will continue refining the proposal as we go. You can see our exact words on this here. It’s totally normal to work on proposals this way.
…
Anyways, hopefully by now you get where I’m coming from. This post has so many flaws that people should weigh its words very carefully. And every minute that went into writing this could have been used improving the Aave Will Win proposal, helping promote Aave, discussing marketing plans with a partner, etc. Even worse, now Aave’s competition has fantastical and untrue propaganda they can use to tarnish Aave’s name both publicly and privately.