Editorial and AI Policy

Tests are completed before a draft is written. The intended editorial order is: define the claim, run the test, preserve the evidence, draft the explanation, check it against the evidence, and publish only after human review.

AI assistance

AI tools may assist with outlining, wording, summarization, consistency checks, or identifying questions for review. AI output is not treated as test evidence and does not replace running the workflow. A human editor remains responsible for the published claims, limitations, sources, and corrections.

We do not knowingly put confidential client data, account secrets, or unnecessary personal information into AI tools for article production.

Limits

Software changes after publication, and neither human nor AI review can guarantee that guidance will remain accurate or work in every environment. Articles state their tested context so readers can judge how closely it matches their own.