[ARFC ADDENDUM] Mandatory Disclosures and Conflict-of-Interest Voting Norms

Governance General
1

Title: [ARFC ADDENDUM] Mandatory Disclosures and Conflict-of-Interest Voting Norms

Author: ACI (Aave Chan Initiative)

Date: 2026-02-03

Summary

This ARFC addendum proposes a governance-framework update to strengthen transparency and conflict-of-interest (COI) norms across Aave governance.

It introduces:

  • Mandatory disclosures for any participant currently receiving, or who has previously received, compensation from the Aave DAO, as well as any candidate seeking compensation.
  • A clear statement that COI voting restrictions cannot be enforced onchain, and should instead be treated as a social-contract norm.
  • A universal expectation of ethical abstention on matters where a voter has a material COI.

Motivation

Aave governance relies on credible neutrality, informed participation, and trust in process. As the DAO grows, the number of compensated contributors and service providers increases, which is healthy. However, without clear, consistent disclosure and COI norms, governance can drift into perceived capture or legitimacy debates that harm the DAO, the protocol, and the $AAVE token.

Valuation models apply a discount to uncertainty. Clear disclosure and COI norms reduce that uncertainty by improving transparency, accountability, and the perceived legitimacy of outcomes.

Specification

1. Mandatory disclosure requirements

Disclosure is mandatory for:

  • Any individual or entity with an active compensation stream from the Aave DAO.
  • Any individual or entity that has received compensation from the Aave DAO in the past.
  • Any individual or entity applying to be compensated by the Aave DAO, including as an author, co-author, or beneficiary of a proposal.

Disclosures must be provided in:

  1. the forum profile disclosure field (delegate profile), and/or
  2. a dedicated service provider presentation thread (or equivalent standardized location), and
  3. an explicit “Disclosure” section in any ARFC, TEMP CHECK, Snapshot, or AIP the disclosing party authors, co-authors, or materially benefits from.

Minimum disclosure content:

  • Entity or role (delegate, service provider, contributor)
  • Nature of compensation (stream, grant, retroactive, consulting)
  • Status (active or past)
  • Relevant scope (workstream or mandate)

Address disclosure (voting power only)

Any entity or individual must publicly disclose all addresses under their control that hold Aave voting power or receive delegated Aave voting power.

For privacy and safety reasons, addresses that hold $AAVE but do not carry voting power (i.e., all voting power delegated to a third-party address) are out of scope.

Recommended standardized disclosure sentence:

“Disclosure: I currently receive, or have previously received, compensation from the Aave DAO via [stream/grant/contract], related to [scope]. Voting power addresses under my control: [0x…], [0x…].”

2. COI voting restrictions are social, not onchain

COI voting restrictions cannot be reliably enforced at the onchain voting level. Any attempt to encode broad COI restrictions at the protocol layer is likely to introduce complexity, edge cases, and enforcement ambiguity that creates more issues than it solves.

Therefore, COI voting restrictions should be treated as a social contract rule:

  • Delegates and voters are expected to follow ethical abstention norms.
  • The community is encouraged to treat votes cast under material COI as lacking legitimacy from an ethical standpoint, even though they remain technically valid onchain.
  • The community, delegates and voters are invited to use their voting power to counteract restrictions violation

3. Ethical abstention due to COI should be universal

Ethical abstention is a baseline governance norm. It applies universally whenever a voter has a material conflict of interest, regardless of topic, author, or outcome.

A material conflict of interest includes, but is not limited to:

  • Any direct or indirect financial exposure that could be meaningfully affected by the vote (streams, grants, fees, revenue share, equity, token arrangements, or contingent compensation).
  • Any employment, advisory role, mandate, or contractual relationship with an entity materially impacted by the decision.
  • Any situation where the voter receives a non-trivial private benefit from the outcome that is not broadly shared by $AAVE token holders.

Expectation:

If a voter has a material conflict of interest, they MUST not cast a YAE or NAY vote, ABSTAIN votes cast are tolerated.

If they do vote YAE or NAY anyway, that vote does not count under this framework. It must be treated as invalid for legitimacy purposes and excluded from any community-recognized “clean” tally, quorum or outcome assessment, even if excluding it would change the result.

For AIPs, if a voter has a material conflict of interest, they MUST not cast a vote.

4. Proposal power should not be restricted

This addendum does not propose restricting proposal power.

Rationale:

  • Proposal power is necessary for executing work, coordinating upgrades, and progressing governance.
  • The neutrality concern is primarily about voting outcomes under COI, not about enabling proposals to be drafted and discussed or created.
  • Restricting proposal power would increase operational friction and reduce execution throughput.

5. Implementation and enforcement approach

Implementation is procedural and forum-based:

  • Improve “Disclosure” expectation to the governance posting templates for ARFCs, TEMP CHECKs, and Snapshot threads.
  • Encourage delegates and service providers to maintain up-to-date profile disclosures.
  • Moderation must remind participants of missing disclosures.

Disclaimer

ACI is a service provider to the Aave DAO and is compensated for its mandates. This addendum is presented in the interest of improving governance legitimacy, transparency, and durable alignment.

Next Steps

  1. Publish a standard ARFC, collect community & service providers feedback before escalating proposal to ARFC Addendum Snapshot,
  2. If the ARFC snapshot outcome is YAE, proposal will be canon.
17 Likes
2

A much-needed proposal to address current voting loopholes. We’ll need a more comprehensive governance constitution eventually, but this is desperately needed right now.

7 Likes
3

Thanks for putting together this proposal, @ACI.

Rules have to evolve in order to better adapt to decentralized governance reality, especially when it comes to transparency and ethics.

4 Likes
4

Thanks for bringing this proposal, as transparency is everything in this industry. Aave (and lots of protocols as well) has everything to gain by adopting these standards

Fully supporting this :+1:

5 Likes
5

I worry that this proposal would strengthen transparency but, in practice, it would introduce confusion, subjectivity, and power asymmetries that are harmful for Aave governance.

For example, declaring onchain-valid votes as “socially invalid” creates a parallel governance system with no rules, no finality, and no clear authority. At Aave’s scale, governance outcomes must be deterministic and final. Otherwise every vote becomes contestable forever.

The definition of “material conflict of interest” is a bit vague and unenforceable. There are no thresholds, no neutral arbiter, and no clear process.

In a mature DAO, almost every important decision materially affects all service providers. In practice, this framework implies that, on the most critical decisions for the protocol, only small token holders with no direct involvement, limited context, and no touch with the Aave business would be able to vote.

There is also a serious asymmetry around delegation. The proposal explicitly excludes delegators from disclosure requirements, while focusing only on voting addresses. In practice, this pushes large holders or whales who wish to remain private toward delegation by default.

It’s my view that this goes directly against core principles of permissionless, self-sovereign participation. Principles that Aave, together with other leading actors of the ecosystem, should continue to uphold and represent.

This structure also runs the risk of disproportionately benefiting delegation aggregators, while the true origin of voting power remains opaque by design. Under the addendum’s own logic, this is a clear and material private benefit for some and not for others. Applying the same COI standard consistently would imply that such actors should not vote on this proposal.

More broadly, the framework proposed is not crypto-native.
Aave is a crypto DAO.

Legitimacy should come from transparent, onchain rules and finality, not from offchain moral judgments or social legitimacy policing.

Many service providers have historically received DAO compensation while acting as major voters, including on proposals where they were allocating funds to themselves. This was never an issue or seriously contested by the DAO. It was accepted, transparent, and is one of the reasons why Aave governance has worked and why we are here today.

Introducing complexity and ambiguity at this stage would only increase friction in Aave’s business and ecosystem relationships and ultimately favour competitors with simpler or no governance models.

For these reasons, I believe this addendum should not move forward.

5 Likes
6

Thank you for the perspective shared in this thread. As we transition from the experimental phase of DAOs into the era of institutional-grade decentralized finance, perhaps we should reflect on the fundamental nature of our governance.

I would like to offer a few questions for the community to consider, drawing a parallel to the evolution we are seeing in other major ecosystems, specifically the Sky Atlas framework:

• On Credible Neutrality: If the primary value proposition of a protocol like Aave is its ‘credible neutrality,’ how do we protect that perception when those tasked with the protocol’s development also hold the power to vote on their own mandates or compensation? Can a system truly be considered decentralized if the ‘observers’ and the ‘actors’ are the same group?

• On the ‘Alignment’ Principle: The Sky Atlas (the foundational constitution of the Sky ecosystem) prioritizes the concept of Alignment. It suggests that participants in the core ecosystem must adhere to strict transparency to prevent ‘Governance Capture.’ If a contributor has an economic interest in a specific outcome, does their participation in the vote strengthen the protocol’s security, or does it introduce a hidden risk premium for other token holders?

• On Social Contracts vs. Technical Constraints: While it is true that on-chain enforcement of ‘intent’ is difficult, does that negate the necessity of a Social Contract? Should Aave’s standards be any less rigorous as we aim for the next trillion in TVL?

The goal of this addendum isn’t to restrict expertise, but rather to ensure that expertise is deployed in a way that is beyond reproach. If we seek to be the global liquidity layer, must we not first ensure our governance processes are as transparent and predictable as our smart contracts?

3 Likes
7

Replying in a personal capacity here, not on behalf of Aave Labs.

Admittedly, I am very worried about creating an additional governance layer that is ruled by social agreement because it seems high risk at the scale Aave operates.

My main concern is that it’s a system that is guaranteed to be inconsistent. It relies on the community to monitor, identify, and punish infractions, which means enforcement is not based on an objective standard (i.e. an onchain/Snapshot vote outcome), but on who is paying attention and who can generate the most outrage.

Some additional questions I have:

  • Who is responsible for creating and publishing the clean tally?
  • What is the process for deciding which votes to exclude? Is there an appeals process for someone whose vote is excluded?
  • Since the onchain vote is what actually matters, how would the clean tally be used to influence governance outcomes after the fact? Does that remove finality from onchain votes?

I do agree that disclosing COIs upfront should be a part of the code of conduct, even in some cases where it may be obvious. But a social judgement system outside of voting is a slippery slope.

3 Likes
8

The risk of ‘social agreement’ at scale is that it often devolves into inconsistent, outrage-based enforcement. If Aave is to function as a global financial utility, we need to move past ‘reputation’ and toward verifiable neutrality.

To address the concerns regarding subjectivity and finality, I’d like to pose a few Socratic questions on how we might institutionalize this addendum:

• On the ‘Who Decides’ Problem: What if we removed the ‘mob’ from the equation entirely? Could we delegate the validation of conflict-of-interest disputes to a neutral, third-party arbitration protocol like Kleros? By using an incentivized jury system, we replace subjective forum sentiment with an objective ruling based on the rules of our Social Contract. Does moving the dispute to a decentralized ‘court’ provide the predictable standard necessary for institutional confidence?

• On Administrative Oversight: Would the creation of a dedicated ‘Governance Council’ (or specialized Stewards) streamline this process? This body wouldn’t have the power to change votes, but rather the mandate to maintain a ‘Disclosure Registry’ and flag potential omissions for review. Could a specialized council provide the consistency that a broad, distracted community cannot?

• On Finality vs. Legitimacy: We should be clear that a ‘Social Contract’ does not break the code; it defines the Legitimacy of the Mandate. In the Sky Atlas, ‘Alignment’ is a prerequisite for ecosystem participation. If an on-chain vote passes but is later ruled ‘conflicted’ by a neutral jury (Kleros), it doesn’t revert the transaction, but it does provide a formal basis for the DAO to initiate a ‘Correction AIP’ or suspend the offender’s service provider status. Isn’t this ‘Social Slashing’ a more robust deterrent than mere reputational damage?

The goal isn’t to create more ‘vibe-based’ governance, but to build a framework where the cost of dishonesty is mathematically and socially prohibitive. If we can trust oracles for our price feeds, why wouldn’t we trust a decentralized jury for our governance integrity?

2 Likes
9

I agree that the intent behind the proposal is reasonable, but any approach like this creates more problems than it solves. Weakening on-chain enforcement undermines the core value of blockchain governance and replaces clear rules with subjective judgment.

I also agree that DAOs are not neutral and are influenced by large holders. This issue exists as a design flaw of most DAOs. But treating on-chain voting as non-canonical and open to disputes only makes the situation worse by breaking the core principles of blockchain decision-making. It also introduces significant bureaucracy and operational friction around any controversial vote

Transparency is better than pretending otherwise. If governance needs improvement, it should focus on making discussions and influence more transparent, especially at the forum level where decisions are shaped. Addressing bots and sybil accounts would be more constructive than reducing the guarantees provided by on-chain voting

2 Likes
10

Simo and ACI bring valuable perspectives, but we are drifting towards a deadlock between “permissionless purity” and “social enforcement.”

Simo is correct: declaring on-chain valid votes as “socially invalid” introduces dangerous ambiguity and threatens finality. Aave is a financial protocol, not a social club; it requires deterministic outcomes.

However, ACI is also correct: in a DAO of this scale, relying solely on “reputation” to curb malicious behavior by massive stakeholders is naive. A rational actor with a conflict of interest (COI) will happily trade social capital for financial gain if the profit exceeds the reputational cost.

In DeFi, we do not govern behavior with “police” or “social pressure.” We govern it with incentives and slashing conditions.

Therefore, I propose moving this discussion away from “Banning” or “Social Shaming” toward a DeFi-native solution: The Bonded Integrity Framework.

Here is how we solve this without breaking the permissionless nature of the protocol:

1. From “Trust” to “Bonding” (Skin in the Game)

Instead of just asking for disclosures, we should introduce a bonding requirement. Any Service Provider (SP) receiving a stream from the DAO must lock a percentage of their compensation (in AAVE or stkAAVE) into a “Integrity Bond” smart contract.

If you want to extract value from the DAO, you must stake value in the DAO.

2. Optimistic Compliance & Slashing (Not Censorship)

To address Simo’s concern: We do not restrict voting wallets.

Every token holder, including SPs with a COI, retains the technical ability to vote. There is no censorship, and on-chain finality is preserved.

However, if an SP votes on a proposal where they have a defined Material Conflict of Interest (e.g., voting to increase their own budget or control), their Bond becomes subject to a Slashing Condition.

3. The Challenge Mechanism (Resolution Layer)

Instead of a “social consensus” deciding validity after the fact, we use an Optimistic Challenge Mechanism:

If an SP violates the COI mandate, community members can trigger a “Challenge” (by posting a collateral).

This triggers a dispute resolution process (via a neutral Arbitration Oracle like Kleros or a specific DAO Ratification Vote).

The Outcome: If the breach is verified, the vote stands (preserving finality), but the SP’s Integrity Bond is slashed and returned to the Treasury.

This is the Industry Standard for Infrastructure

This proposal is not experimental; it aligns Aave with the most robust security models in the ecosystem:

EigenLayer secures the Ethereum ecosystem not by asking operators to “be nice,” but by enforcing Slashing Conditions on staked assets.

Arbitrum & Optimism protect billions in TVL using Optimistic Challenge Mechanisms (Fraud Proofs) to verify honesty without slowing down the chain.

MakerDAO (Endgame) is moving toward “Aligned Delegates” who face economic consequences for breaking their scope.

Why This Works for Aave:

This transforms the issue from a moral debate into a mathematical calculation.

If an SP wants to push a conflicted vote, they are free to do so but they must be willing to lose their bond. If the cost of the slash exceeds the benefit of the vote, they will rationally choose to abstain.

4 Likes
11

Thanks for putting this forward, I think the core direction is right and timely. Clear disclosure requirements are exactly what we need to reduce the current uncertainty and restore governance legitimacy. From a token holder perspective and a DAO member, better transparency is not “bureaucracy”, it’s risk management and it directly impacts how markets price $AAVE.

That said, I agree this should likely evolve into a V2 to avoid shifting the problem into an unstructured “soft governance” layer. In particular, the “clean tally / social invalidity” concept is useful as a norm, but without a clearly defined process it risks becoming subjective, contested, and ultimately another battleground.

I’m supportive of the paths suggested by @ApuMallku and @Gross as the natural next step:

  • some form of institutionalized dispute resolution / integrity process (e.g., a lightweight governance integrity council or an arbitration layer for challenges), and/or

  • a more DeFi-native bonded framework where mandate holders/service providers put skin in the game and face enforceable consequences for material COI violations (e.g., funding/mandate implications, or bonding + challenge mechanisms).

Practically, I’d be in favor of adopting this addendum now as a baseline (mandatory disclosure + universal COI abstention expectation), while explicitly committing to a follow-up V2 that defines, who adjudicates COI disputes, what the appeal path is, and how enforcement ties back to mandates/funding rather than relying purely on social pressure.

Overall, strong step in the right direction for the DAO’s interests clarity first, then a tighter enforcement mechanism so legitimacy doesn’t depend on vibes.

5 Likes
13

Reading through this addendum, it resonates with situations that surface when you actually try to follow delegated execution and value routing step by step.

The mechanics can be correct, but without a shared understanding of roles and boundaries during execution, it becomes difficult for participants to reason about responsibility in practice. From that perspective, this feels like a constructive move toward improving not just formal structure, but the common ground needed for cooperation and informed participation.

Systems like this only really work when understanding can circulate as easily as value, and it’s good to see that being acknowledged and supported here.

1 Like
14

I support the transparency updates but I would suggest conflicted voters cast an ABSTAIN vote instead of not voting at all.

Silence is ambiguous. It creates uncertainty about whether a delegate is actually ethically ‘compliant’, negligent, or simply absent.

Voting ABSTAIN puts clear intent onchain: it shows the delegate is active, aware of the conflict, and deliberately choosing not to influence the outcome, making compliance verifiable.

And could even comment on the vote.

This approach is consistent with practices in other DAOs. For example in Arbitrum, the Code of Conduct requires mandatory COI disclosure but does not forbid voting or require non-participation, allowing delegates to remain active while managing conflicts transparently.

Similar in Uniswap, the Delegate Code of Conduct explicitly advises delegates to abstain when they cannot fully evaluate a proposal, and recognizes disclosure plus abstention as an appropriate response in cases of conflict or uncertainty.

4 Likes
15

Hello,

Thanks for the feedback on this proposal

  1. @Ignas agree on abstain, but not as a mandatory vote, COI voter can either vote that or not cast a vote at all

  2. not surprised at Labs taking their approach and strategy, no comment needed, if think everyone has formed their own opinion now

  3. I echo several feedbacks and @Emereb summarizes it best, we need a quick patch now to mitigate the “slow motion coup” the protocol is currently experiencing and it’s worth after this proposal to take time to implement something more future-proof.

2 Likes
16

I was naive thinking they’d embrace community consensus on transparency, resilience, and decentralization—all for the product’s benefit. Multiple community members have noted that several protocols already have this type of “social contract” as standard practice.

Why should we lower our standards? Aren’t we aiming for “trillions”?

Labs’ complete tone-deafness is heartbreaking.

4 Likes
17

This feels a bit like bait, but to clarify: I personally am on board with COI disclosures.

My only question is about how a new social governance layer will be executed and if that makes finality of onchain voting more difficult.

1 Like
18

We’ve already provided substantial feedback and referenced examples from other successful protocols. Could you check with your Labs colleagues to see if they have alternative approaches? If there’s consensus that COI disclosures would benefit the protocol, let’s collaborate on making Aave stronger.
Wouldn’t productive collaboration be healthier than continued isolation?
I believe it would be far more constructive if Labs proposed solutions rather than dismissing concerns with flippant comments or ignoring legitimate questions—which has unfortunately been the pattern. As a significant token holder, I feel completely unheard by Labs. The continued ghosting is deeply frustrating.
That’s why, together with @phenk53, we developed this proposal: [TEMP CHECK] Aave Service Provider & Orbit Delegate Revenue Alignment Framework

5 Likes
21

We should draw on the U.S. political system.

Most DAOs are lacking in populism and are controlled by a small number of oligarchs and whales.

The House of Representatives is established, adopting a square-root weighted voting mechanism. Under this system, the increase in voting weight slows as holdings grow, thereby appropriately diluting the relative advantage of large holders and enhancing the influence of ordinary retail participants. This ensures governance better reflects the genuine will of the broader community.

Simultaneously, the Senate is established, with participation eligibility restricted to users who have staked their assets for one year or longer. This represents the interests of committed, long-term holders deeply engaged in the ecosystem, safeguarding the protocol’s long-term stability and sustainable development.

22

The current proposal has been escalated to ARFC Addendum Snapshot.

Vote will start tomorrow, we encourage everyone to participate.

2 Likes
23

We hope that Aave Labs can promptly release a draft of the relevant proposal for thorough community discussion. The matter has been delayed for over a month, with no draft or text released to the public. Without a foundational document or clear framework, the community is unable to engage in effective communication, negotiation, or compromise, and is forced to initiate proposals independently to drive progress.

As a core team with long-term and deep involvement in protocol development, Aave Labs should take the initiative in drafting proposals, facilitating communication and coordination, and fostering consensus on such critical matters involving significant stakeholder interests, the future direction of the protocol, and community trust. Prolonged silence, with no drafts, no framework, and no timeline, is not an acceptable approach.

Only by presenting a formal draft can the community engage in rational discussions, make mutual concessions, and seek optimal compromises. This is essential to truly restore trust, avoid internal conflicts, and allow Aave to focus on preparing for the next phase of industry growth. We urge the team to act promptly and provide a clear, discussable, and actionable draft proposal for voting.