Using AI for Marking and Feedback: A Guide for Australian Teachers

Marking is one of the most time-consuming parts of teaching, and one of the most important. Quality feedback is a significant driver of student learning — research consistently identifies it as one of the highest-impact teaching practices. The challenge is that providing quality feedback at scale is genuinely difficult, particularly for teachers managing large classes.

AI tools can help — but the way they are used for marking and feedback requires more care than other teaching applications, because of the privacy implications of working with student work.

---

The Privacy Consideration

Before using any AI tool for marking or feedback, Australian teachers need to understand the privacy implications.

Student work is personal information. Under the Privacy Act 1988 and state and territory privacy legislation, schools have obligations about how they collect, use, and disclose student information. Entering identifiable student work — work that includes a student's name, student number, or other identifying information — into a consumer AI tool raises privacy concerns that most school privacy policies do not permit.

The practical guidance from most Australian state and territory education departments is:

• Do not enter identifiable student work into consumer AI tools (ChatGPT, Claude, etc.)
• Use AI tools that operate within your school's existing data agreements (Microsoft Copilot within Microsoft 365, Google Gemini within Google Workspace for Education)
• When using consumer AI tools for feedback-related tasks, describe the work in general terms rather than pasting the actual student work

---

What AI Can Do for Marking and Feedback

Developing Rubrics

One of the most time-consuming parts of assessment is developing a good rubric. AI tools can generate detailed rubrics for a wide range of assessment tasks in minutes. A well-designed rubric makes marking faster and more consistent, and provides students with clear expectations before they begin the task.

Example prompt:

Create a detailed marking rubric for a Year 9 English persuasive essay. The essay should be 600-800 words and address a contemporary issue. The rubric should assess: (1) argument and evidence, (2) persuasive language techniques, (3) text structure and organisation, and (4) grammar and expression. Use 4 performance levels: Excellent, Proficient, Developing, and Beginning. Align the descriptors to the Australian Curriculum v9.0 Year 9 English achievement standard.

Writing Feedback Language

AI tools can help teachers develop the language for written feedback — particularly for teachers who find it difficult to write feedback that is specific, constructive, and encouraging at the same time.

The approach that protects student privacy is to describe the work in general terms rather than pasting it:

Example prompt:

Help me write feedback for a Year 9 student whose persuasive essay demonstrates a clear argument and good use of rhetorical questions, but lacks specific evidence to support the main claims and has inconsistent paragraph structure. The feedback should acknowledge the strengths, identify the key area for improvement, and suggest a specific next step. Keep the tone encouraging and specific. Aim for 3-4 sentences.

Generating Whole-Class Feedback

After marking a set of assessments, AI tools can help teachers write whole-class feedback that addresses common patterns.

Example prompt:

I have just marked 28 Year 9 persuasive essays. Common strengths were: clear thesis statements, effective use of rhetorical questions, and confident voice. Common areas for improvement were: lack of specific evidence, weak conclusions that simply repeated the introduction, and inconsistent use of formal language. Write a whole-class feedback statement I can share with students that acknowledges the strengths, addresses the common issues, and provides specific guidance for improvement.

Writing Report Comments

School report comments are one of the most time-consuming writing tasks teachers face each semester. AI tools can generate draft comments based on a brief description of the student's performance.

Example prompt:

Write a school report comment for a Year 9 English student who is performing at the Proficient level. The student demonstrates strong analytical skills when discussing texts, contributes thoughtfully to class discussions, and produces well-structured written responses. The main area for development is building confidence in extended writing tasks under timed conditions. The comment should be 3-4 sentences, use positive and constructive language, and avoid generic phrases.

---

Workflow for AI-Assisted Marking

Here is a practical workflow for using AI to support marking and feedback while protecting student privacy:

Step 1: Develop the rubric before marking begins. Use AI to generate a detailed rubric, then review and adjust it. A good rubric makes the marking itself faster and more consistent.

Step 2: Mark against the rubric. Do the actual marking yourself, using the rubric. Note the key strengths and areas for improvement for each student.

Step 3: Use AI to help write feedback language. For students whose feedback is difficult to write, describe the work in general terms and ask the AI to help draft feedback language. Review and personalise the output.

Step 4: Generate whole-class feedback. After marking the full set, use AI to help write whole-class feedback based on the patterns you observed.

Step 5: Write report comments. Use AI to generate draft report comments based on brief descriptions of each student's performance. Review and personalise each comment before submitting.

---

Using AI Within Your School's Data Agreements

For teachers whose schools use Microsoft 365 or Google Workspace for Education, the AI tools integrated into those platforms offer a privacy-safer option for some marking and feedback tasks.

Microsoft Copilot in Word: If you have student work in Word documents within your school's Microsoft 365 environment, Copilot can help you draft feedback comments directly within the document.

Google Gemini in Docs: Similarly, if student work is in Google Docs within your school's Google Workspace for Education environment, Gemini can assist with feedback within that environment.

Before using these features with student work, confirm with your school's IT coordinator that your school's agreement covers this use case.

---

Maintaining Professional Standards

Using AI for marking and feedback does not reduce the teacher's professional responsibility for the quality of that feedback. The AITSL Australian Professional Standards for Teachers require teachers to provide feedback that is timely, specific, and useful for student learning — regardless of what tools are used to produce it.

This means:

• Review every piece of AI-generated feedback before it reaches a student
• Personalise feedback to reflect the specific student and their work
• Ensure feedback is accurate — AI tools can make errors, particularly in subject-specific content
• Apply professional judgment about what feedback is appropriate for each student's stage of development

---

Getting Started

The most practical starting point is rubric development. Next time you are designing an assessment task, use an AI tool to generate a draft rubric before you start marking. Review and adjust it, then use it for marking. The time saved in marking — because you have a clear, detailed rubric to work from — will be immediately apparent.

From there, experiment with using AI to help draft whole-class feedback after your next marking round. This is a low-risk application that does not involve student data and can save significant time.