Executive Order
Knowledge Integrity & AI-Verification Standards Protocol
Preamble
WHEREAS, Aevum Encyclopedia has expanded to over 2.4 million verified articles across 140 languages, supported by an active contributor base of 180,000 scholars, researchers, and subject-matter experts; and
WHEREAS, the integration of generative artificial intelligence into research, drafting, and cross-referencing workflows has accelerated content production while introducing new challenges regarding hallucination mitigation, citation transparency, and editorial accountability; and
NOW, THEREFORE, by the authority vested in the Office of the Chief Knowledge Officer and in accordance with the Aevum Governance Charter (Rev. 3.1), the following directives are hereby issued to standardize AI-assisted knowledge production, ensure traceable sourcing, and maintain academic-grade integrity across all platform content.
Purpose & Scope
This Executive Order establishes mandatory protocols for AI-assisted drafting, real-time verification, and contributor tier classification. It applies to all editorial staff, contracted researchers, automated pipelines, and community contributors publishing under the Aevum Encyclopedia namespace.
Directives
AI-Assisted Drafting Limitations
All AI-generated text must be explicitly tagged at the paragraph level using the [AI-Assist: Verified] metadata flag. No article may exceed 40% AI-generated content without mandatory dual-review by Level-3 or higher human editors.
Citation & Provenance Requirements
Every factual claim, statistic, or historical assertion must link to at least two independent, publicly accessible primary or peer-reviewed secondary sources. AI verification engines will automatically flag unanchored claims for human audit before publication.
Contributor Tier & Access Matrix
Contributors are hereby classified into four tiers based on verification history, domain expertise, and peer endorsement:
- Tier 1 (Observer): Read-only, suggestion queue access
- Tier 2 (Drafting): Minor edits, citation additions, requires 24h review buffer
- Tier 3 (Editor): Full drafting rights, cross-disciplinary tagging, AI tool access
- Tier 4 (Domain Lead): Final approval authority, policy override within scope, audit access
Transparency Dashboard & Public Audit Log
A real-time public dashboard will be deployed at aevum.org/transparency, displaying verification success rates, AI-assist ratios, editorial turnaround times, and resolved dispute metrics. All editorial actions will be logged immutably and accessible for external academic audit.
Implementation Timeline
- June 1–15, 2025: System migration & contributor notification period
- June 15, 2025: Full enforcement of metadata tagging & citation engines
- July 1, 2025: Tier classification migration complete; legacy accounts grandfathered
- August 1, 2025: Public Transparency Dashboard v1.0 launch
Compliance & Enforcement
Non-compliance with these directives will result in content suspension, contributor tier demotion, or permanent access revocation depending on severity. The Editorial Oversight Committee retains final adjudication authority. Appeals may be submitted via the official governance portal within 14 business days of action.
Severability
If any provision of this Order is found to be invalid or unenforceable, the remaining provisions shall continue in full force and effect. Amendments require a supermajority vote of the Governance Board and public comment period of no less than 30 days.