Keep AI-written hints in identifiable records separate from exercise questions and solutions. Document how each component was created, which version it supports, and what remains in the final book. A human-written exercise does not make a generated hint human-written, and a correct answer does not prove that the hint gives an appropriate amount of help.
Separate the components of an exercise
Give each exercise a stable identifier and distinguish its prompt, data or source material, hints, answer, and explanation. These components may come from different people or processes. Treating the entire exercise as one undifferentiated block makes its history harder to understand and its revisions harder to coordinate.
Record which components appear in the student-facing pages and which appear later or in a separate edition. A hint intended for an optional support section should not accidentally enter the main prompt during a merge. The component record should preserve both origin and intended placement.
Define the help each hint should provide
Write the hint's instructional purpose before generating alternatives. One hint might identify a useful concept, while a later hint might suggest the first step. Decide how much of the solution should remain for the learner to discover and keep that boundary in the brief.
Avoid equating a shorter hint with a better hint. A few words can reveal the answer immediately, while a longer explanation can still leave the central reasoning to the learner. Review the hint according to the task it supports rather than a fixed word-count target.
Record generation and revision decisions
When a tool drafts hints, retain the relevant request, accepted result, and the exercise version used as context. Record later human changes and rejected alternatives where they explain the final choice. Keep that history separate from the claim that the finished exercise has been technically checked.
Use the current KDP distinction when preparing the publication record. Generated text remains generated material even after substantial edits; human-created text improved through AI follows a different creation path. Document what actually happened to each component rather than applying one convenient label to the entire exercise set.
Review hints against the current solution
Compare every hint with the approved solution and the knowledge the reader is expected to have. A hint may point toward an outdated method after the exercise is revised, or require a concept that the chapter has not introduced. Those problems can remain even when the hint is grammatically clear.
Test graduated hints in sequence and independently where readers can access them out of order. Confirm that each one gives the intended level of support and that the final hint does not contradict an earlier one. Keep the result of this review attached to the actual accepted wording.
Preserve the relationships during manuscript changes
If the question, numerical values, illustration, or expected answer changes, flag its hints for review. Do not assume that a hint remains valid because it avoids mentioning the changed value. It may still rely on a method or relationship that the revised exercise no longer uses.
Track copies across student pages, answer sections, instructor notes, and companion material included in the project. Retire old hints explicitly so a later export or resumed generation job does not restore them from an earlier draft. Stable component identifiers make that cleanup more reliable than searching for one remembered sentence.
Reconcile the final book with the component records
Before export, confirm which hints actually remain and where readers will encounter them. Check that optional help has not become an accidental spoiler through page placement, headings, or navigation. Then compare the displayed text with the reviewed component version.
Use the final inventory to support accurate creation records and publication disclosures. Keep instructional review and origin documentation distinct: one addresses whether the hint helps the learner, while the other explains how the content was produced. Both should remain available when the exercise is revised for another edition.
Frequently asked questions
Clear answers for this publishing decision.
Does a human-written answer change the origin of a generated hint?
No. Preserve the creation history of each component. The fact that one part was written by a person does not establish how the other parts were produced.
Can I reuse one hint across several similar exercises?
Review its usefulness and accuracy for each exercise, and record those uses. Similar-looking questions can require different reasoning or different amounts of support.
What if the AI only suggests an idea for a hint?
Record whether you used generated wording or wrote the actual hint yourself from the idea. That distinction helps describe the creation process accurately under the current guidance.
Should hints be checked again after layout?
Check their placement, visibility, and final wording. A hint can be instructionally sound in the source but reveal too much when placed beside the question or linked in an unintended way.
What must I check, disclose, and verify myself when AI helped produce the book?
Review facts, rights, and source material yourself, record which parts were AI-assisted or AI-generated so you can answer Amazon’s disclosure question accurately, and treat every automated check as a starting point rather than a verdict.
See what the compliance checks cover