Can You Use AI for HSC Assessments? A Practical Checklist

A student finishes a Year 12 Physics depth study report. The explanation is clear, the structure is polished, and the references look impressive. Then the teacher asks, “Show me how you developed this section.”

The student opens a chat history and realises they cannot explain why the AI chose those ideas, which parts are correct, or where the information came from.

That is the key boundary.

Using AI to learn is different from using AI to produce work that is supposed to show your own knowledge. The decision you need to make is not “Is AI good or bad?” The useful question is:

“Is AI helping me build my understanding, or is it replacing the thinking my assessment is meant to measure?”

NESA’s rules require HSC assessment marks to reflect a student’s own knowledge, skills and understanding. Unapproved use of AI in completing assessment tasks can breach academic integrity rules, and schools must provide their own policies and instructions about acceptable use. NSW Government

01The quick AI decision rule

Before using AI for an assessment, ask three questions:

1. What is the purpose of this task?
   If the task is designed to test your explanation, argument, calculations, creativity, or problem-solving, be careful about letting AI do those parts.

2. What does my assessment notification allow?
   Your teacher’s instructions are the first place to check. Different tasks can have different rules.

3. Could I explain and defend every part of what I submit?
   If the answer is no, the work probably does not represent your own learning.

A simple rule:

The word “usually” matters. Schools create their own policies within NESA requirements, so never assume a use is allowed just because another student, teacher, or website says it is. NSW Government

02The three zones: study, permission, and risk

A useful way to think about AI is not “allowed” versus “banned”. Think about the level of control you still have over the learning.

03Zone 1: AI uses that are generally suitable for independent study

These uses are about practising, checking, and understanding.

The AI is acting like a study partner. It can explain ideas, but it cannot sit the exam for you.

Examples

Explaining a difficult concept

You are revising Chemistry and do not understand why increasing temperature can change equilibrium position.

A useful prompt:

“Explain Le Chatelier’s principle using a simple example, then give me a harder HSC-style explanation question.”

You still need to decide whether the explanation makes sense and connect it to your syllabus.

Generating practice questions

You can ask AI:

“Create five Year 12 Mathematics Extension 1 questions on integration by substitution. Do not give solutions yet.”

Then attempt them yourself.

This is similar to creating your own flashcards or extra practice questions.

Getting feedback on your understanding

You could write:

“Here is my explanation of electric fields. Find any physics errors and ask me questions that test my understanding.”

The important part is that your explanation existed before the AI feedback.

This type of active learning is usually more valuable than simply rereading notes. If you want to improve your revision approach, see Why Rereading Notes Feels Better Than It Works.

04Zone 2: Uses that need your teacher’s explicit permission

These uses sit in the middle because they may support learning, but they can also affect the final assessment product.

The same AI action can be fine in one task and not allowed in another.

Example: improving a draft

Imagine you write an English Extension essay introduction yourself.

You ask AI:

“Does my argument clearly answer the question? Give feedback using the marking criteria.”

This is different from:

“Rewrite this introduction so it sounds like a Band 6 response.”

The first asks for feedback. The second asks AI to create a stronger version of your assessed writing.

Whether editing, restructuring, or feedback is permitted depends on the assessment conditions.

Example: research support

You might ask AI to:

• suggest search terms,
• explain unfamiliar terminology,
• help organise research notes.

But you still need to check sources, decide what evidence matters, and write the final explanation yourself.

AI can confidently produce incorrect information. A polished answer is not automatically a correct answer.

Example: coding or data analysis

For a Mathematics or Science task, asking AI:

“Explain why this graph has this shape”

is very different from:

“Create the entire analysis section for my investigation.”

The second removes the reasoning the task may be designed to assess.

05Zone 3: Actions that risk presenting AI work as your own

These are the situations where students often get into trouble.

Copying generated paragraphs into a submission

The paragraph might be accurate. It might even sound like your writing.

The problem is that the assessment is no longer showing your ability to construct that response.

Asking AI to solve the whole task

Examples:

• “Write my Physics depth study report.”
• “Complete my Chemistry research response.”
• “Make my Mathematics investigation.”
• “Answer this take-home assessment.”

The final product may look complete, but the learning evidence belongs to the AI process, not your own work.

Using AI output without acknowledging it when required

NESA expects students to acknowledge sources and materials appropriately, including generative AI where relevant. NSW Government

06How to check before you use AI for an assessment

Do not rely on guesses. Use a checking process.

Step 1: Read the assessment notification carefully

Look for statements about:

• whether AI is allowed,
• whether AI is completely prohibited,
• whether AI can be used for planning, research, editing, or feedback,
• whether AI use must be acknowledged.

The assessment notification is often the most specific instruction you will receive.

Step 2: Read your school’s malpractice policy

Schools must provide students with their assessment and malpractice procedures. These policies explain what your school considers unacceptable behaviour. NSW Curriculum

A school policy may include details about:

• acceptable AI tools,
• acknowledgement requirements,
• record keeping,
• consequences of unauthorised use.

Step 3: Ask a precise question if you are unsure

Avoid asking:

“Can I use AI?”

Ask something specific:

“Can I use AI to check grammar in my draft before submission?”

or:

“Can I use AI to generate practice questions while preparing, if none of the questions appear in my final response?”

Specific questions are easier for your teacher to answer clearly.

07Keeping evidence that your work is genuinely yours

Good evidence is not about proving you never used technology. It is about showing your thinking process.

Useful records include:

• handwritten planning pages,
• draft versions,
• research notes,
• calculation attempts,
• experiment records,
• teacher feedback,
• version history of documents.

For example, in a Physics investigation, a strong evidence trail might show:

1. Your initial hypothesis.
2. Your first calculations.
3. Problems you found.
4. Changes you made after checking your method.
5. Your final explanation.

The final report is only one snapshot. The process shows how you got there.

10Common mistakes when using AI for HSC work

Mistake 1: “If AI gives me the right answer, it is fine.”

Why it happens:
AI answers often look confident and polished.

Why it fails:
Assessment tasks measure your ability to reason, communicate, and apply knowledge. A correct answer from AI may not show that you can do those things.

Better approach:
Use AI to test your understanding after you attempt the problem.

Mistake 2: “Everyone uses AI, so the rules must be flexible.”

Why it happens:
Students hear different advice from friends and online sources.

Why it fails:
Your school’s assessment instructions apply to your task.

Better approach:
Check the notification and ask your teacher.

Mistake 3: “I only changed a few words, so it is still my work.”

Why it happens:
Editing can feel minor.

Why it fails:
The important question is not how many words changed. It is whether the AI contributed the thinking being assessed.

Better approach:
Keep your original drafts and be clear about what help you received.

Mistake 4: “AI detection will decide whether I get caught.”

Why it happens:
Students often hear about AI detectors online.

Why it fails:
NESA states that schools should not rely on AI detection software as the main safeguard when determining malpractice. The focus is on evidence of authentic student work and school procedures. NSW Curriculum

Better approach:
Focus on producing work you can explain and support.

11Questions and solutions

Question 1

You are studying Mathematics Extension 1. You ask AI to generate five integration questions, complete them yourself, and then use AI to check your solutions. Is this generally a suitable study use?

Solution 1

Yes, this is generally a suitable study use because AI is supporting practice rather than producing your assessed work.

The student created the mathematical reasoning by attempting the questions. The AI is acting as a checking tool.

A good habit is to compare the AI explanation with your own working. If the AI says your answer is wrong, you should verify the mathematics yourself rather than automatically accepting the correction.

Question 2

A student uses AI to create a first draft of a Biology depth study report. They then rewrite some sentences and add their own references. Is this enough to make it their own work?

Solution 2

No, rewriting some sentences is not automatically enough because the AI created the underlying structure and explanation.

The tempting argument is that the student “edited” the work, but the important constraint is what the assessment is measuring. If the task assesses the student’s ability to research, analyse, and communicate scientific ideas, those parts need to come from the student unless AI use is specifically allowed.

The student should instead create their own draft and use permitted support, such as feedback on clarity or checking understanding.

Question 3

A student is completing a take-home Mathematics assessment. The notification says AI may be used for “brainstorming ideas” but not for producing solutions. The student asks AI for three possible methods, chooses one, and completes all calculations independently. Is this acceptable?

Solution 3

It may be acceptable because the student has followed the stated boundary, assuming the brainstorming use is exactly what the notification permits.

The important step is separating idea generation from solution creation.

The student still needs to:

• perform the calculations,
• justify the chosen method,
• check the final answer,
• acknowledge AI use if the instructions require it.

If the AI had produced the full solution and the student copied the method without understanding it, the situation would be different.

Question 4

A Chemistry student uses AI to summarise a scientific article. They then write their own report using the summary but never read the original article. What is the main problem?

Solution 4

The main problem is that the student has allowed AI to become a substitute for understanding the source material.

The tempting route is efficient: a summary saves time. However, scientific reports depend on accurate interpretation of evidence.

The student should check the original article, confirm important claims, and make their own decisions about which evidence supports their argument.

Question 5

A student completes a major investigation. During the project, they use AI to suggest possible experimental variables, explain unfamiliar terms, and comment on drafts. They keep all their drafts and notes. The final submission is their own work, and the assessment instructions allow these AI uses. Is this the same as submitting AI-generated work?

Solution 5

No, these situations are different because the student remains responsible for the reasoning and final product.

The tempting assumption is that “any AI involvement means it is not authentic”. That is too simple. The key question is whether AI replaced the skills being assessed.

In this case:

• the student selected the variables,
• designed the investigation,
• interpreted the results,
• created the final explanation,
• kept evidence of their process.

The constraint that changes the answer is the purpose of the AI use. Support that strengthens learning is different from AI completing the learning task on the student’s behalf.

12The next decision to make

AI is becoming another tool in the HSC environment, like calculators, reference books, and online resources. The skill is not avoiding every tool. It is knowing when the tool helps you think and when it starts thinking for you.

Before your next assessment, do not start with “What can AI do for this task?”

Start with:

“What part of this task is my teacher actually trying to see me do?”

That answer usually reveals the boundary.