ZIP-603: Research DAO Governance

Abstract

This ZIP defines the governance framework for Zoo's Research DAO, a purpose-built decentralized autonomous organization for allocating research funding. The DAO uses quadratic voting to amplify the voice of individual community members over large token holders, milestone-based grant disbursement to ensure accountability, and an on-chain conflict of interest detection system to prevent self-dealing. The Research DAO operates within the legal structure of Zoo Labs Foundation (501(c)(3)) as described in ZIP-0 Sections 10-18, with the Foundation board retaining a narrow compliance veto.

Motivation

Research funding allocation has historically been controlled by small committees of grant program officers at agencies like NSF, NIH, and the European Research Council. While peer review provides quality control, this centralized model has well-documented biases:

Conservatism: Panels favor incremental research over high-risk, high-reward proposals. Breakthrough ideas are systematically underfunded.
Network effects: Researchers with existing institutional connections and track records receive disproportionate funding. Early-career researchers and those from underrepresented institutions are disadvantaged.
Geographic bias: Funding is concentrated in wealthy nations. Conservation research in biodiversity-rich developing nations is chronically underfunded despite having the greatest impact potential.
Slow cycle times: Traditional grant cycles take 6-18 months from submission to funding decision. Conservation emergencies (disease outbreaks, habitat destruction events) cannot wait.
Opacity: Funding decisions are opaque. Rejected applicants receive minimal feedback. There is no public record of why one proposal was funded over another.

Zoo Labs Foundation requires a funding mechanism that is transparent, community-driven, bias-resistant, and fast enough to respond to conservation emergencies. The Research DAO provides this mechanism while remaining compliant with 501(c)(3) requirements for charitable fund deployment.

Cross-Ecosystem Context

LP-8800 through LP-8805 (Lux DAO Governance Platform): On-chain proposal lifecycle, voting mechanics, and treasury integration patterns.
LP-8993 (Lux Community Development Grants): Milestone-based grant disbursement and community evaluation framework.
HIP-0066 (Hanzo Data Governance): Conflict of interest detection and data access control patterns.

Specification

1. DAO Structure

1.1 Governance Hierarchy

Zoo Labs Foundation Board (501(c)(3) compliance veto)
    |
    v
Research DAO Council (elected, 7-11 members)
    |
    +-- Funding Committee (reviews proposals, recommends allocation)
    +-- Ethics Committee (reviews COI, ensures ZIP-540 compliance)
    +-- Technical Committee (reviews methodology, feasibility)
    |
    v
ZOO Token Holders (vote on proposals, elect council)
    |
    v
KEEPER Token Holders (donor governance weight, non-transferable)

1.2 Council Election

Parameter	Value
Council size	7 members (expandable to 11 by governance vote)
Term length	6 months
Election method	Ranked-choice voting, ZOO + KEEPER weighted
Eligibility	1000+ ZOO staked for 90+ days OR 500+ KEEPER
Minimum voter turnout	10% of circulating ZOO
Seat reserve	At least 2 seats for contributors from Global South institutions

1.3 Council Responsibilities

Responsibility	Description
Proposal triage	Assign proposals to review committees
Reviewer assignment	Select peer reviewers per ZIP-600 Section 2
Emergency funding	Authorize emergency grants up to 5,000 ZOO without full vote
Parameter adjustment	Propose changes to governance parameters
Dispute resolution	Final arbiter for funding disputes
Compliance reporting	Quarterly report to Foundation board

2. Proposal Process

2.1 Proposal Types

Type	Funding Range	Approval Mechanism	Timeline
Micro-grant	1-500 ZOO	Council approval (simple majority)	7 days
Standard grant	501-10,000 ZOO	Quadratic vote + council ratification	30 days
Major grant	10,001-100,000 ZOO	Quadratic vote + council supermajority + board non-veto	60 days
Emergency grant	1-5,000 ZOO	Council emergency approval (5/7 minimum)	48 hours
Retroactive reward	1-50,000 ZOO	Quadratic vote	30 days

2.2 Proposal Submission

Proposals follow the ZIP-600 schema with additional governance fields:

{
  "proposalId": "bytes32",
  "proposalType": "enum(MICRO, STANDARD, MAJOR, EMERGENCY, RETROACTIVE)",
  "title": "string",
  "abstract": "string",
  "fullProposalCid": "string",
  "principalInvestigator": "address",
  "fundingRequested": "uint256",
  "milestones": [...],
  "conflictDisclosure": {
    "affiliations": ["string"],
    "priorFunding": ["string"],
    "relationships": ["string"],
    "financialInterests": ["string"]
  },
  "impactStatement": {
    "conservationOutcome": "string",
    "communityBenefit": "string",
    "openScienceCommitment": "string"
  }
}

2.3 Proposal Lifecycle

DRAFT
  |
  v
SUBMITTED (bond deposited per ZIP-600)
  |
  v
TRIAGE (council assigns to committee, 3 days max)
  |
  v
REVIEW (peer review per ZIP-600 Section 2, committee evaluation)
  |
  v
COI_CHECK (automated + manual conflict screening, see Section 4)
  |
  v
VOTING (quadratic vote for Standard/Major/Retroactive)
  |
  v
RATIFICATION (council ratification for Standard; council + board non-veto for Major)
  |
  v
FUNDED (escrow created, first milestone unlocked)
  |
  v
ACTIVE (milestones tracked per ZIP-600 Section 3)
  |
  v
COMPLETED (final report submitted, remaining funds released)

3. Quadratic Voting

3.1 Voting Mechanism

Quadratic voting (QV) ensures that the cost of additional votes increases quadratically, preventing whale dominance:

VotingPower = sqrt(tokensAllocated)

Example:
  1 ZOO   = 1.0 vote
  4 ZOO   = 2.0 votes
  9 ZOO   = 3.0 votes
  100 ZOO = 10.0 votes
  10000 ZOO = 100.0 votes

3.2 Voice Credits

Each voting epoch, participants receive voice credits based on their governance weight:

VoiceCredits = BaseCredits + StakingBonus + KeeperBonus

Where:
  BaseCredits   = 100 (equal for all participants)
  StakingBonus  = min(100, sqrt(staked_ZOO))
  KeeperBonus   = min(50, sqrt(KEEPER_balance))

Voice credits are non-transferable and expire at the end of each voting epoch (30 days). Unused credits do not carry over.

3.3 Vote Options

Option	Description
YES	Fund the proposal
NO	Do not fund the proposal
ABSTAIN	Counted for quorum, not for outcome

3.4 Approval Thresholds

Proposal Type	Quorum (% of active voice credits)	Approval (YES / (YES + NO))
Standard	15%	> 50%
Major	25%	> 60%
Retroactive	15%	> 50%

3.5 Matching Funds (Quadratic Funding)

For Standard and Major grants, community YES votes trigger matching funds from the Research Treasury:

MatchingAmount = (sum(sqrt(individual_contribution_i)))^2 - sum(individual_contribution_i)

This is the standard quadratic funding formula (Buterin, Hitzig, Weyl 2018). The matching pool is capped at the treasury allocation for the current epoch.

4. Conflict of Interest Detection

4.1 Automated Screening

The protocol performs automated COI screening at proposal submission and before each vote:

Check	Method	Action
Self-funding	PI address matches voter address	Auto-recusal from vote
Institutional affiliation	DID credential matching	Flag, require disclosure
Co-authorship history	On-chain publication graph (ZIP-600)	Flag if co-authored within 3 years
Prior funding relationship	Proposal history matching	Flag if PI reviewed voter's proposal
Token transfer history	On-chain transfer graph analysis	Flag if direct transfer within 90 days
Shared multi-sig membership	Gnosis Safe co-signer detection	Flag, require disclosure

4.2 COI Severity Levels

Level	Description	Action
NONE	No detected conflicts	Normal voting
MINOR	Indirect relationship (same institution, no direct collaboration)	Must disclose, may vote
MODERATE	Direct collaboration history, shared funding	Must disclose, vote weight halved
SEVERE	Financial interest, direct relationship with PI	Mandatory recusal
CRITICAL	Self-dealing or undisclosed severe conflict	Recusal + reputation penalty

4.3 COI Appeals

Contributors flagged for COI may appeal:

Submit written explanation to Ethics Committee.
Ethics Committee rules within 7 days.
Appeal to full Council if Ethics Committee ruling is disputed.
Council decision is final.

4.4 Smart Contract Interface

interface IConflictDetector {
    function checkConflict(address voter, bytes32 proposalId) external view returns (
        uint8 severityLevel,
        string[] memory reasons
    );
    function declareConflict(bytes32 proposalId, string calldata disclosure) external;
    function recuse(bytes32 proposalId) external;
    function appealConflict(bytes32 proposalId, string calldata justification) external;
}

5. Milestone-Based Disbursement

5.1 Milestone Evaluation

Each milestone is evaluated by the Technical Committee:

Criterion	Weight	Description
Deliverable completion	40%	All stated deliverables produced
Data quality	20%	Data meets ZIP-601 quality standards
Open access compliance	15%	All outputs published as committed
Timeline adherence	10%	Delivered within deadline (grace period: 14 days)
Budget compliance	15%	Spending within 20% of budgeted amounts

Score >= 60%: Milestone approved, funds released. Score 40-59%: Conditional approval, remediation plan required within 14 days. Score < 40%: Milestone failed, remediation period per ZIP-600.

5.2 Disbursement Schedule

Proposal funded (total: N ZOO)
    |
    v
Milestone 1 (X% of N): Released upon M1 approval
    |
    v
Milestone 2 (Y% of N): Released upon M2 approval
    |
    ...
    v
Final milestone (Z% of N): Released upon completion + final report
    |
    v
Holdback (10% of N): Released 90 days after completion if no disputes

The 10% holdback incentivizes researchers to remain responsive for post-completion questions and data requests.

6. Treasury Management

6.1 Research Treasury

The Research Treasury is a dedicated allocation from the Zoo Treasury:

Allocation	Source	Governance
Core research fund	Zoo Treasury allocation (quarterly)	DAO vote per ZIP-0 Section 9
Quadratic matching pool	Dedicated matching fund	Sized per epoch by Council
Emergency fund	5% of core fund, ring-fenced	Council access only
Retroactive fund	10% of core fund	DAO vote for retroactive awards

6.2 Fund Accounting

All disbursements are tracked on-chain with per-proposal accounting:

interface IResearchTreasury {
    function getEpochBudget() external view returns (uint256);
    function getMatchingPool() external view returns (uint256);
    function getEmergencyFund() external view returns (uint256);
    function getProposalDisbursed(bytes32 proposalId) external view returns (uint256);
    function getEpochDisbursed() external view returns (uint256);
}

7. 501(c)(3) Compliance

7.1 Board Veto

Per ZIP-0 Section 15, the Foundation board retains a narrow compliance veto:

Proposal approved by DAO
    |
    v
Board review (Major grants only, 7-day window)
    |
    v
Board action:
  - NO ACTION (default): Proposal proceeds
  - VETO (with published reason): Only for law/mission/donor-restriction violation

Veto reasons are recorded on-chain and must cite specific legal or policy grounds. The board cannot veto based on scientific disagreement.

7.2 Charitable Purpose Alignment

All funded proposals must advance Zoo Labs Foundation's charitable purposes:

Conservation science and wildlife protection
Open AI research for environmental benefit
Environmental education and community science
Decentralized science infrastructure as public good

8. Protocol Parameters

Parameter	Default	Governance-Adjustable
Council size	7	Yes (7-11)
Council term	6 months	Yes (3-12 months)
Base voice credits	100	Yes
Voting epoch	30 days	Yes
Micro-grant ceiling	500 ZOO	Yes
Standard grant ceiling	10,000 ZOO	Yes
Emergency grant ceiling	5,000 ZOO	Yes
Milestone holdback	10%	Yes (5-20%)
Holdback release delay	90 days	Yes
Board veto window	7 days	Yes (3-14 days)

Rationale

Quadratic voting: QV is chosen over simple token-weighted voting because research funding decisions should reflect community breadth, not capital concentration. A whale should not be able to single-handedly fund a project that benefits their own portfolio.
Council layer: Pure direct democracy is too slow and lacks domain expertise for research evaluation. The elected Council provides fast triage and emergency response while community votes control major decisions.
Mandatory COI detection: Self-dealing is the primary governance attack vector in DAO-funded research. Automated on-chain detection catches obvious conflicts; the Ethics Committee handles nuanced cases.
Milestone disbursement: Adapted from LP-8993 (Lux Community Development Grants). Milestone-based release with holdback is standard practice in both traditional grants and crypto ecosystem grants (Optimism RPGF, Gitcoin).
501(c)(3) board veto: The narrow veto preserves the legal structure required for tax-deductible donations while minimizing centralized control. The board cannot override community decisions on scientific grounds.

Security Considerations

Governance attacks: A well-funded attacker could acquire ZOO tokens to influence voting. Quadratic voting significantly increases the cost of such attacks (cost grows quadratically with desired influence). The Council ratification layer provides a human check.
Collusion: Voters could coordinate to concentrate quadratic funding matching on a specific proposal. The COI detection system flags unusual voting patterns and shared financial relationships. Anti-collusion mechanisms from MACI (Minimal Anti-Collusion Infrastructure) may be adopted in a future ZIP.
Council capture: Council members could collude to approve self-serving micro-grants that bypass community voting. The 5,000 ZOO emergency ceiling, public on-chain records, and Foundation board oversight mitigate this risk.
Vote buying: Off-chain vote buying (paying someone to vote YES) is difficult to detect. Future integration with MACI would make vote buying provably ineffective by allowing voters to change votes privately after the buying agreement.
Treasury drain: A sustained attack of many small proposals could drain the treasury without triggering Major grant oversight. Epoch-based budgets cap total disbursement per period. The Council monitors aggregate spending.
Flash loan voting: ZOO acquired via flash loan could be used to vote. The protocol requires tokens to be staked for the full voting period (30 days minimum), making flash loan attacks economically infeasible.

References

ZIP-0: Zoo Ecosystem Architecture & Framework (Sections 9-18)
ZIP-540: Research Ethics & Data Governance
ZIP-600: DeSci Protocol Framework
LP-8800 through LP-8805: Lux DAO Governance Platform
LP-8993: Lux Community Development Grants
HIP-0066: Hanzo Data Governance
Buterin, V., Hitzig, Z., Weyl, E. G. "A Flexible Design for Funding Public Goods." Management Science 65(11), 5171-5187 (2019).
Optimism Collective. "Retroactive Public Goods Funding." https://community.optimism.io/docs/citizen-house/
Gitcoin. "Quadratic Funding." https://wtfisqf.com/

Changelog

Version	Date	Changes
1.0	2025-01-15	Initial draft

Research DAO Governance