S4-S6 History: AI evidence-trail workshop (three classroom tasks)

• StageS4-S6
• SubjectHistory
• Learning strandreflection on the application of AI
• Official documentsAI Literacy Learning Framework for Primary and Secondary Schools · Guide to Using AI in Teaching in Primary and Secondary Schools
• Session time1 lesson (60 min)

An AI answer in history often arrives as a fluent paragraph with no sources, and students cannot tell a sourced claim from a plausible one. Three linked tasks teach S4-S6 students to split a claim, trace the evidence, and rewrite it into a version they can stand behind, with a reusable evidence-trail record, a worked example and a marking guide.

• AudienceTeachers Students
• ThemeStudent thinking

• EquipmentPrinting or projection; for one step pupils may need the school's approved AI tool
• Version1.8
• Last updated2026-10-17

What this solves

The hardest thing in senior-secondary history is often not whether pupils use AI, but that the AI’s answer comes back as a fluent paragraph with no provenance. Pupils cannot tell which sentence is backed by evidence and which only sounds right; reading through a chat history is slow, and it does not show whether the pupil actually made a judgement.

This workshop makes provenance something pupils can see and write down: three linked tasks take one AI history answer and split the claim, trace the evidence, then rewrite it into a version the pupil can stand behind. The teacher can then see where each pupil’s thinking stops — whether they cannot separate fact from interpretation, do not know where to look for a source, or dare not cut a claim that has no trail.

The stimulus uses a famous, textbook-level historical episode, and is deliberately written as a “can this claim stand” judgement rather than a task that asks pupils to verify a date or figure no one can check. Every school and example is fictional.

The workshop can sit inside a school’s digital education planning and be run in a senior-secondary history lesson, assessed through process evidence rather than learning data alone; the fluent but unsourced answer it works on is exactly the shape an AI hallucination takes, and pupils learn to verify each claim and fact-check it against a source.

Learning objectives

By the end, pupils can:

1. Split an AI history statement sentence by sentence into “fact claim” and “interpretation”, and mark every place a source is missing.
2. For two claims they judge traceable, say what kind of primary or secondary source would settle each one and where in Hong Kong they would look for it (for example the school library, Hong Kong Public Libraries, the Government Records Service, or a museum archive).
3. Rewrite the AI paragraph into a version they can stand behind, deleting or clearly hedging every claim with no trail.

It targets the reflection on the application of AI strand of AI literacy and builds critical thinking about Technology for Good.

How to use it

One 60-minute lesson, in four segments.

1. Introduction, ground rules and splitting the claim (15 min) — Set the goal: “We are not judging whether AI should be used; we are learning to trace where a claim comes from.” Agree the ground rule: these tasks do not grade whether a pupil used AI. Hand out the sheet and the evidence-trail record. Pupils read a short AI-generated history statement (fictional) and label each sentence “fact claim” or “interpretation”, marking every place a source is missing.
2. Find the trail (20 min) — Pupils choose two claims they judge traceable and write what kind of source would settle each one and where in Hong Kong they would look. The teacher runs through what each kind of archive or library is for, without naming a specific document.
3. Defend or drop it (18 min) — Pupils rewrite the AI paragraph into a version they can stand behind: keep what has a trail, delete or hedge what does not, and account for each item on the record.
4. Wrap-up and follow-up (7 min) — Choose one follow-up activity (below) and collect the record as evidence for what to teach next.

Answer key and marking notes

All three tasks are open-ended, but good answers share checkable features. The examples below are fictional; teachers can swap in a famous episode from their own course.

Stimulus (a fictional AI answer)

Hong Kong grew from a small fishing village into an important trading port (entrepot) during the nineteenth century. This change happened mainly because the British brought a policy of free trade, so Hong Kong’s economy took off at once. Most historians agree that Hong Kong’s prosperity was caused by the advantage of its location. We can therefore say that, without British rule, Hong Kong would have remained a small fishing village for ever.

The evidence-trail record (for pupils)

Columns: claim | fact or interpretation | source that would settle it | kept / hedged / dropped.

Worked example (completed)

Marking guide

Task 1: Split the claim

Task 2: Find the trail

Task 3: Defend or drop it

What not to grade

Do not grade whether a pupil used AI. This workshop scores evidence work — splitting a claim, saying what kind of source would settle it, cutting or hedging what has no trail — not the tool a pupil used. Using AI is not a fault here; the point is to see where the pupil’s thinking stops and what to teach next.

Follow-up

• Have pupils run their last history essay’s concluding paragraph through the same record, sentence by sentence.
• Turn the class’s list of “where in Hong Kong to look” into a poster of which kind of source to seek and where.

Files

A printable worksheet (with the stimulus, the evidence-trail record, answer spaces, the marking guide and a worked example) is in the download file:

Worksheet (Word document, editable)

All examples and schools are fictional and do not involve any real school, pupil or document.

Official documents

Every resource in this library is written to the AI Literacy Learning Framework for Primary and Secondary Schools and falls within the policy scope of the Blueprint for Digital Education Development in Primary and Secondary Schools.

This resource maps to:

• Supplement I: AI Literacy Learning Framework for Primary and Secondary Schools (p.1-13): calls for assessment that values application, ethics and thinking processes, rather than memorising the final answer alone.
• Supplement II: Guide to Using AI in Teaching in Primary and Secondary Schools (pp.2-7 to 2-8): recommends checking pupils’ learning and AI use through oral questioning and the submission of thinking processes, rather than collecting finished work alone.

Learning strand: reflection on the application of AI

Official documents page: https://www.edb.gov.hk/en/edu-system/primary-secondary/applicable-to-primary-secondary/it-in-edu/debp.html

This page is an independent Beets teaching resource, not an official EDB document.

Feedback

If you have any comments or questions about this resource, or an idea for the next one, tell us on the feedback page.