For civil-society advocates

1 · Where the catalog is most useful to advocacy today

Topics with the strongest cross-jurisdiction coverage that advocacy coalitions cite most often:

• Biometric identification

  Carve-outs for law enforcement are where the EU AIA Article 5 prohibition gets contested. The article surfaces both sides.

• Deepfakes / synthetic content

  EU AIA Art. 50 + multiple US state laws + India MEITY advisory. Election-integrity overlap.

• AI in employment

  Hiring algorithms, workplace surveillance. Gig-worker algorithmic management is a deferred sub-topic — see Section 3.

• AI in criminal justice

  Risk-assessment, sentencing assistance, predictive policing. Documented harms section is a known gap.

2 · How to use the catalog in an open letter or regulatory submission

• Find the topic articlecovering your campaign's subject matter (e.g., /wiki/biometric-id).
• Pull the citationusing the "Cite this" widget at the bottom of the article (BibTeX, APA, Chicago, Harvard, OSCOLA, Bluebook, RIS available).
• Cite by version using the ?asOf=YYYY-MM-DD permanent URL. This guarantees the version of the article you cited will be retrievable indefinitely, even if the live catalog later updates.
• For comparative arguments(e.g., "the EU prohibits this, China requires it") use the comparative jurisdictional memos at /wiki/compare. Each memo carries a PDF download with a stable URL.
• Embed the article card in an explainer post using the iframe at /embed/wiki/[slug]. CSP allows embedding from any origin.

3 · Known gaps (deliberate transparency)

The civil-society-persona audit run in iter-312 surfaced these gaps. Each is tracked publicly so we can't silently decide we've addressed them:

• Gig-economy algorithmic management is not yet a distinct topic. Coverage is folded into /wiki/employment, which under-serves gig-worker advocacy. Splitting is on the roadmap.
• Public biometric surveillance is currently folded into /wiki/biometric-id. The civil-society lens distinguishes capability (biometric matching) from deployment (real-time surveillance in public spaces). Splitting is on the roadmap.
• Automated welfare denial is absent. Not yet covered by any catalog row. This is a high-priority gap for the next topic-proposer round.
• Immigration AI / biometric detection is absent. Not yet covered.
• Documented harms + affected-communities content is absent from article templates. Adding this requires coalition partnership — we won't author harm-narrative content without affected-community input. See Section 4.
• Power-asymmetry annotationon coverage cells (e.g., "EU AIA Art. 5 prohibits, but Art. 26(10) carves out law enforcement — the carve-out shifts the effective rule") is not yet a structured field. Roadmap.
• Global South paradigm parity: catalog currently frames Global South instruments as implementing OECD paradigms rather than as rival governance philosophies. Promoting development-rights and tech-sovereignty topics to the decision-support-header tier is on the roadmap.

4 · How to propose a coverage gap

We accept structured proposals from civil-society organisations and affected-community networks. The fastest path is a GitHub issue using the template below. If your organisation doesn't use GitHub, email advocates@policywindow.org with the same information and we'll file the issue on your behalf with attribution.

Template (paste into a GitHub issue):

## Proposed coverage gap

**Submitting organisation:** [name + link]
**Affected community:** [who is harmed / served by addressing this]
**Harm class:** [discrimination / surveillance / displacement / denial of service / other]

**Jurisdictions where this surfaces:**
- [jurisdiction] — [instrument or absence of instrument]
- [jurisdiction] — [instrument or absence of instrument]

**Primary sources to cite:**
1. [URL — operative text or regulatory guidance]
2. [URL — case study or affected-community report]

**Suggested catalog action:**
- [ ] Add a new topic: [code + label proposal]
- [ ] Add coverage cells for [instrument] × [topic]
- [ ] Add a debate (rival positions on this topic)
- [ ] Update existing topic description: [current vs proposed]
- [ ] Other: [describe]

**Why now:**
[one paragraph: regulatory momentum, upcoming consultation, recent harm event]

**Editorial-merit checklist** (the editorial board uses this to
evaluate; honest answers improve the proposal's chance):
- [ ] Primary sources exist and are accessible
- [ ] The proposed topic / cell is a stable category, not a
  fast-moving news event
- [ ] At least 2 jurisdictions are tracked or addressable
- [ ] The proposing organisation has subject-matter expertise on
  the affected community (this is not a credibility test — we
  just want to know who you are)

Issues go to the public repository. The proposer's identity is visible in the change record. Decision target: 30 days for an initial editorial response; longer for proposals that need primary-source verification or new topic-kind work.

5 · How we handle competing interpretations

Many topics on the catalog are contested — civil-society and industry read the same regulatory text differently. The /wiki/debates catalogue is the structured surface for this: each debate page lays out the competing positions with primary-source citations for each side. Topic articles link out to relevant debates via the "Related debates — rival interpretations & counterevidence" section.

If you think a debate page is missing a position or unfairly framing one, file an issue using the template in §4 above. We don't take sides — but we do want every position represented accurately so a reader can evaluate.