Rules for accountable publishing
Editorial policy
ai-summary.net serves a US-first audience of students and self-directed learners. Pages should answer a real learning question clearly, show the evidence behind material claims, and distinguish product facts from guidance and opinion.
Source hierarchy
- Primary sources first: official platform and provider documentation, standards, original research, and current app evidence.
- High-quality secondary sources when needed: used to explain context, never to override a more direct current source.
- Search results and AI output are discovery aids: they are not evidence by themselves and cannot be cited as proof.
Claim rules
- Attribute provider and platform behavior to the named source.
- Use current source code or a reproducible current-build check for app behavior; label inferences as inferences.
- State meaningful limitations next to the workflow or claim they qualify, not in a distant disclaimer.
- Do not publish invented search volume, rank, accuracy, speed, credentials, awards, endorsements, or review claims.
- Do not imply that a summary replaces the original source or professional judgment.
Required disclosures
AI assistance disclosure
AI tools may assist with research, outlining, drafting, and editing. Heni Hazbay remains responsible for checking sources, product behavior, limitations, and the final page before it can be published.
Creator-interest disclosure
Heni Hazbay created Summarise Visually and has a direct interest in the app. Product references on this site come from that position and should not be read as independent recommendations.
Human accountability before publication
A technically complete draft remains a draft until Heni checks and approves its material claims, sources, current product presentation, limitations, and final wording. AI assistance cannot approve a page or take accountability for publication. The site must not say that Heni reviewed a page until that review has occurred.
Corrections
Send a correction report to [email protected]. Include the page URL, the statement you believe is wrong, and a primary source when one is available. Do not send confidential source material, passwords, payment details, or private screenshots. No response-time promise is made.
When reproducible evidence of an error reaches the publisher through an approved channel, the affected page returns to draft review. A material correction updates the claim, supporting source, limitation where relevant, and review date after verification.
See the research and review methodology for the full evidence workflow and About Heni Hazbayfor the creator-interest context.