Editorial & Verification Policy

This policy describes the editorial principles NAPR intends to use when documenting organizational AI policies.

Primary-source preference

NAPR should prefer first-party policy pages, official statements, corporate publications, terms, charters, or other primary evidence whenever reasonably available. Reliable secondary reporting may provide context or help locate a primary source.

Classification is evidence-based

NAPR classifications should reflect what a source supports—not what NAPR assumes an organization probably does. A limited restriction should not be promoted to a comprehensive NAPR-0 pledge merely because an organization expresses skepticism about AI.

Verification date

Published policy records should identify when NAPR last verified the supporting source. Verification means the source was reviewed for the recorded claim; it does not mean NAPR audited every internal practice of the organization.

Policy versions and history

When a material policy change is identified, NAPR should preserve or record the prior state rather than silently overwriting history. Current and superseded positions should be distinguishable.

Conflicts and uncertainty

If sources conflict or a policy’s scope is genuinely ambiguous, NAPR should record the uncertainty, seek clarification where appropriate, or withhold a stronger classification until evidence supports it.

Commercial independence

Payment, advertising, sponsorship, verification fees, consulting relationships, or other commercial arrangements should never purchase a favorable classification. If NAPR later charges for a verification process, payment should purchase review effort—not a predetermined outcome.

Corrections

Material factual errors should be corrected. Significant classification disputes or substantive record changes should be documented when practical. See Corrections & Disputes.

AI assistance

NAPR may use AI-assisted research and editorial tools subject to the separate AI Use & Editorial Transparency policy.