Core Principles
All edits, whether suggested by AI-assisted tools or human contributors, must adhere to these foundational standards. Deviations will be flagged for review.
⚖️ Neutral Point of View (NPOV)
Articles must fairly represent all significant, published perspectives on a topic. Biased language, promotional tone, or unverified claims are prohibited.
🔍 Verifiability & Sourcing
Every factual claim must be traceable to a reputable, primary or secondary source. AI-generated summaries must cite their training references. Uncited material may be removed.
🚫 No Original Research
Aevum does not publish unpublished data, personal theories, or novel synthesis. Content must reflect established, peer-reviewed, or widely recognized knowledge.
🤝 Civility & Respect
Discussions on talk pages and review queues must remain constructive. Harassment, trolling, or bad-faith editing results in immediate suspension.
©️ Copyright & Media Licensing
All uploaded media must be public domain, Creative Commons, or properly licensed. AI-generated images must be clearly tagged. Plagiarism is strictly enforced.
Editing Workflow
Every change passes through a structured pipeline to balance rapid knowledge updates with rigorous verification.
Draft
Contributor submits edit via web or mobile editor.
AI Pre-Check
Automated validation for citations, tone, and policy compliance.
Peer Review
Subject-matter experts or verified editors assess accuracy.
Publication
Changes go live with full version history and attribution.
Turnaround time: Minor edits are typically reviewed within 2–24 hours. Major rewrites or sensitive topics may take up to 72 hours.
Content Standards
Structure & Formatting
- Use clear headings (H2–H4) with logical hierarchy
- Lead paragraph must define the topic concisely (1–3 sentences)
- Tables and infographics must include source attribution
- Mathematical formulas and code blocks require syntax highlighting
Citation Requirements
Accepted sources include peer-reviewed journals, academic publishers, government databases, and reputable news agencies. Personal blogs, unverified forums, and self-published material are not acceptable as primary sources.
Translation & Localization
Multilingual editions must maintain conceptual accuracy. Direct translations are encouraged, but culturally specific examples should be adapted where necessary. All translations undergo native-speaker review.
Contributor Tiers & Trust Levels
Access and editing privileges scale with demonstrated reliability and domain expertise.
Tier 1 Observer
Read-only access. Can suggest edits via the "Propose Change" form. All suggestions require full peer review.
Tier 2 Contributor
Verified account with completed training module. Can edit non-sensitive articles. Changes are fast-tracked through AI validation.
Tier 3 Verified Editor
Demonstrated track record of 50+ accepted edits. Can review other contributors, resolve minor disputes, and edit sensitive topics.
Tier 4 Subject Expert
Professionally credentialed (PhD, industry certification, or institutional affiliation). Has final authority on domain-specific content and mediates complex disputes.
Reporting & Dispute Resolution
If you encounter vandalism, bias, or policy violations, use the "Report Issue" button on any article. Reports are triaged within 24 hours.
Dispute Escalation Path
- Talk Page Discussion: Contributors attempt to resolve differences directly.
- Mediation Request: A neutral verified editor reviews both sides and suggests compromises.
- Policy Committee: Unresolved cases involving policy interpretation are escalated to the Editorial Board for a binding decision.
Repeated or severe policy violations result in temporary suspension or permanent revocation of editing privileges. All enforcement actions are documented and appealable.