Evidence map
Is the harm real, and does governance work?
A black-letter map says what the rules are. This says what the evidence shows. For every evidence-reviewed topic, two social-science questions: is the underlying harm empirically real (problem-reality), and is there rigorous evidence that governing it actually works (governance-efficacy)? Each badge is the epistemic status of the evidence — not the policy debate. Findings are AI-drafted under charter §7.13, epistemic-honesty reviewed, and citation-corroborated; every one is sourced on its topic page. Machine-readable at /wiki/evidence.json.
The second silence
A wiki of rules can map where governance is silent. This layer maps a second silence: 23 of 23 evidence-reviewed topics have thin or absent evidence that the governance actually reduces the harm (100%), and 19 pair a real (or genuinely contested) harm with empirically unproven governance. The case for acting often exists; the replicated case that the chosen rules deliver usually does not.
Coverage: 23 of 23 published topics evidence-reviewed (100%).
Is the harm real?
- established7
- contested12
- thin4
- absent0
Does governance work?
- established0
- contested0
- thin6
- absent17
By topic
| Topic | Harm real? | Governance works? | Signal |
|---|---|---|---|
| Agentic AI Governancecapability | thin | absent | — |
| AI-Driven Worker Displacementsector | contested | thin | real harm · unproven governance |
| Biometric Identificationcapability | established | thin | real harm · unproven governance |
| Catastrophic & Existential Riskcapability | contested | absent | real harm · unproven governance |
| Compute + Model-Weight Export Controlsprocedural | contested | absent | real harm · unproven governance |
| Compute-Threshold Reportingprocedural | contested | absent | real harm · unproven governance |
| AI in Criminal Justicesector | contested | thin | real harm · unproven governance |
| Deepfakes / Synthetic Contentcapability | established | thin | real harm · unproven governance |
| Development-Rights Framingspolitical_frame | thin | absent | — |
| AI in Educationsector | established | thin | real harm · unproven governance |
| AI in Employmentsector | established | absent | real harm · unproven governance |
| Environmental Impact of AI Trainingprocedural | established | absent | real harm · unproven governance |
| Foundation Models / GPAIcapability | contested | absent | real harm · unproven governance |
| AI in Healthcaresector | established | absent | real harm · unproven governance |
| International Coordinationmeta | contested | absent | real harm · unproven governance |
| National Security Carveouts in AI Regulationmeta | thin | absent | — |
| Open-Weight Frontier Releaseprocedural | contested | absent | real harm · unproven governance |
| Individual Redressprocedural | established | absent | real harm · unproven governance |
| Sovereign AI Doctrinepolitical_frame | thin | absent | — |
| Synthetic Content Provenanceprocedural | contested | thin | real harm · unproven governance |
| Technological Sovereigntypolitical_frame | contested | absent | real harm · unproven governance |
| Training-Data Rightsprocedural | contested | absent | real harm · unproven governance |
| Transparency Obligationsprocedural | contested | absent | real harm · unproven governance |
Topic links open the article's sourced evidence section. The slug uses the topic code with underscores as hyphens.