Purpose
AIGovernanceRatings.com publishes institutional research and reference material intended to make AI governance ratings more interpretable, evidence-based and decision-useful. Editorial governance is designed to distinguish factual source material from analytical interpretation and to keep public research separate from issued rating decisions.
Research principles
Traceability
Material factual claims should be attributable to an identifiable source, dataset, methodology document or clearly described analytical basis.
Proportionality
The strength of a claim should reflect the quality, scope, currency and limitations of the supporting evidence.
Separation
Research publication, commercial relationships and rating issuance should remain distinguishable in purpose, governance and disclosure.
Source hierarchy
Primary laws, regulators, standards organizations, government publications, official methodology documents, original datasets and original research are preferred for material claims. Secondary sources may be used for context or discovery, but material analytical claims should be grounded in primary laws, standards, official methodology documents, original datasets or directly attributable research where reasonably available.
Data provenance and citation
Where a publication relies on external data, it should identify the publisher and source sufficiently for a reader to locate the underlying material. Where relevant, the publication should identify the period, jurisdiction, methodology version or access context. Data obtained from third parties remains subject to the source publisher’s definitions, sampling, revisions, corrections and limitations.
Related research from AIGX Research™ may be cited where it supplies relevant frameworks, benchmark design, Responsible AI analysis or sector context. Operative AIGR™ rating methodology should be sourced to AIGR™ Global. Citation of an affiliated source does not remove the obligation to identify external primary sources for material regulatory, standards or factual claims.
Research and rating separation
A public research paper is not an issued AIGR™ rating. Research conclusions should not predetermine the rating of a particular organization, and commercial relationships should not determine public research conclusions or final rating outcomes.
Versioning and review
Research publications should carry a publication reference, version, status and review date. Material revisions to definitions, methodology interpretation, standards mappings or data analysis should be reflected through a version or revision record where the change could affect reader interpretation.
Corrections
Material factual errors should be corrected promptly after verification. If a correction changes a substantive conclusion, the publication should be updated in a way that makes the revision reasonably visible. Correction requests may be submitted through the Contact page.
Conflicts and independence
Authors and reviewers should disclose material interests or relationships that could reasonably affect interpretation. Commercial activity, methodology ownership and related research relationships should be described transparently rather than presented as unrelated third-party independence.
Use of AI-assisted tools
AI-assisted tools may support research organization, drafting, comparison or quality review, but they do not replace accountable human review. Sources, quotations, legal references, quantitative claims and methodology statements should be verified against the underlying material before publication.