AI use, corrections, and human responsibility
AI tools may assist with outlining, language editing, classification, summarization, code support, or structured analysis. Assistance does not transfer editorial responsibility to a model.
Human responsibility
A human reviewer remains responsible for the published claim, source selection, privacy boundary, interpretation, and correction. AI-generated wording must not be treated as evidence.
Verification
Material factual claims require verification against suitable sources or controlled observations. If the available evidence cannot support a claim, the text should narrow it, label the uncertainty, or remove it.
Disclosure
Material AI assistance should be disclosed when it affects how a reasonable reader evaluates authorship, testing, analysis, imagery, or evidence. Routine spelling or formatting assistance does not need to be described as original research.
Privacy and confidential material
Private, client, credential, account, or security-sensitive information must not be placed into an AI service merely because the service is convenient. Use requires an approved privacy and access boundary.
Corrections
Correction requests should identify the page, disputed passage, supporting source, and requested change. Accepted material corrections should be recorded with a date and concise explanation.
No public form, response-time promise, or correction inbox is asserted in this V1 draft. Those details require an approved service and public contact route.
Final CTA: Review the evidence standards → /editorial-policy/
