Loading...
Loading...
Loading…
How we check what we publish, where AI is used and where it stops, and how readers can raise an issue. Written for parents, teachers and school leaders who need to know what stands behind the content their students see.
Last reviewed: 2026-05-15
A. Why verification matters
English revision depends on accurate quotations, clear interpretation, valid exam guidance and trustworthy feedback. A misremembered Macbeth line or a mark-scheme guidance point that is half-right can cost a student real marks in a closed-book exam. Verification is how we keep that surface area clean.
We are an AI-assisted platform, not an AI-only platform. The role of human review is to catch the things AI confidently gets wrong: invented quotations, conflated characters, plausible-sounding but invalid mark-scheme language, and contextual claims that no source supports.
B. Verification principles
C. Review categories
Every published page sits in one or more of these states. Surfacing the labels here keeps the framework honest: a reader can ask which category any specific page is in.
Initial draft, not yet published to the public site.
Drafted with AI assistance and awaiting human review.
A human editor has read the content end-to-end and signed it off.
D. Literature verification
E. Language and writing resources
F. Exam-board alignment
G. AI-generated feedback
H. Correction workflow
A reader spots something that looks wrong and uses the report-a-content-issue link.
The report is recorded with the page URL, the specific claim or quote, and the reporter's notes.
An editor reviews the original source and the claim, plus surrounding pages that reference the same fact.
If the report is valid, the page is updated. If it is a matter of interpretation, we note the alternative reading.
The corrected version replaces the previous content. The page's "Last updated" date is bumped.
For corrections that change the meaning of the page, a short note is added so returning visitors see what changed.
Report it and we will review the page. Reports go to our editorial team, are logged, and acknowledged within one working day.
Mark-scheme guidance, set-text analysis and exam-style examples are reviewed by a human before they reach students. AI assists drafting; humans decide what ships.
Quotations cross-referenced against a reliable source where one is available.
Mapped against public specification information and the relevant assessment objectives.
A reader flagged an issue; the content was reviewed and updated.
No longer in active use. Kept on file for change history; superseded by a newer version.
Claims about a writer's biography, the historical context, or the reception of a text are checked against published criticism. Unverifiable claims are removed or softened.
Students and teachers can challenge or report AI feedback at any point. We treat disputed feedback as a content issue and route it through the correction workflow below.