Editorial-board onboarding

1 · COI disclosure form

Before any editorial work, complete + submit a conflict-of-interest disclosure. The form below is the template; fill it in and email to editorial@policywindow.org (or open a private GitHub issue tagged coi-disclosure).

## Editorial board — Conflict-of-interest disclosure

Editor name:
ORCID:
Affiliation(s):
Position / role:

### Employment + advisory
- Current employer / institution:
- Other board seats, advisory roles, or fellowships:
- Any of the above that involves AI governance, AI policy, or
  regulated AI deployment (yes/no — if yes, name):

### Financial relationships (past 3 years)
- Direct funding / grants from AI labs, regulators, or tech firms:
- Consulting contracts on AI-governance matters:
- Equity stakes (>=1%) in AI labs or AI-regulated companies:
- Speakers' bureau / honoraria from organisations in the catalog:

### Family + personal relationships
- Immediate family in AI labs, AI-governance organisations,
  regulators, or civil-society groups in the catalog:
- Personal relationships that could create a perceived conflict:

### Published positions (intellectual COI ≠ financial COI)
- Public positions you've taken on contested topics in the catalog
  (e.g., position on the EU AI Act, foundation-model regulation,
  algorithmic-discrimination law). List with sources:

### Self-declared recusal triggers
- Topic codes you will recuse from automatically:
- Instrument shortCodes you will recuse from automatically:

### Update commitment
- I will update this disclosure within 7 days of any material change.
- I understand updates are public (logged in the editorial-board git
  history).

Signed:
Date:

The form is reviewed by the project lead (until board ≥3 editors, when review shifts to majority of editors). Standard turnaround: 5 business days. If accepted, you receive onboarding session scheduling + GitHub-team assignment in the same response.

2 · Methodology primer (60 min self-study)

Read in order:

1. /wiki/methodology — the Three Rs framework + catalog-derived rendering + topic-determination
2. /wiki/charter — operating commitments + §7.2 disagreement resolution + §7.3 editorial independence
3. /wiki/reproducibility-policy — why public reproducibility-verdict UI is gated, and what you're committing to when you sign off on one
4. /wiki/coverage-games + the Q2 2026 event record at /wiki/coverage-games/2026-q2
5. /wiki/ai-disclosure — where AI is used in the project; the structural commitments you are upholding by joining

3 · Catalog schema tutorial (45 min)

Read the typed catalog source — that's the entirety of the schema you'll be editing:

• src/lib/international-governance/instruments.ts — GovernanceInstrument + GovernanceTopic interfaces + CoverageType + TopicKind enums + the COVERAGE map
• src/lib/international-governance/instruments-data/ — one file per instrument; read 3 (EU-AIA-2024, US-EO-14110, CN-GENAI-2023) to see the patterns
• src/lib/wiki/concepts.ts — GovernanceConcept rows
• src/lib/capability-evals/benchmarks.ts — Benchmark + SCORE_CLAIMS rows
• src/lib/international-governance/meta-debates.ts — GovernanceDebate rows (rival positions on contested topics)
• src/lib/international-governance/organizations.ts — GovernanceActor rows (regulators, AI labs, civil-society, standards bodies)

4 · Calibration session (2 hours, scheduled with project lead)

A live walk-through covering 5 cells from your topic cluster. For each, the project lead and you independently classify, then compare. The goal is alignment on what high/medium/low confidence means in practice + on the rubric for governs/implicit/silent/conflicts at edge cases.

Schedule via editorial@policywindow.org (Calendly link added once the editorial-board recruitment formalises). Video-call format. Output: 5 calibrated cells + a written shared-understanding summary added to your editorial-board entry under "Calibration notes".

5 · GitHub team + repository access

After COI form acceptance + calibration session:

1. Provide your GitHub handle.
2. You're added to the appropriate CODEOWNERS team for your slot (e.g., @policywindow/wiki-leads for the capability-classes slot).
3. PRs touching catalog files in your topic cluster auto-route to you for review.
4. You're added to a private editorial channel (mailing list or Discord — TBD per board preference) for asynchronous discussion.
5. Your name + ORCID + COI disclosure summary appear on/wiki/editorial-board within 24 hours of all prior steps completing.

6 · Ongoing time commitment (honest estimate)

The 2–4 hours/quarter estimate on /wiki/editorial-board is the median for "quiet" quarters. Realistic range:

7 · What you receive

• Named attribution on /wiki/editorial-board with your ORCID linked — a citable record of editorial service.
• Decision authorityon editorial questions in your topic cluster (at board ≥3: majority vote; at board < 3: founder retains tie-breaker with public disagreement-record per charter §7.2).
• Co-authorship credit on quarterly impact reports (project-level summaries of catalog growth, Coverage Games outcomes, citations received). Credit format: alphabetical author list with editor roles disclosed.
• Recognition pathway when commercial revenue exists — the project is committed to revenue-share with editors once ARR exceeds $100k. Mechanism still being designed in collaboration with the (eventual) full board.

8 · Exit + rotation

Editor service is renewable annually with no maximum term today. To resign: notify editorial@policywindow.org with at least 30 days notice so your topic cluster can be re-staffed. Your past work + named attribution remain in the git history; your editorial-board entry transitions to "former editor (departed YYYY-MM)" rather than being removed.