Got a pile of marking staring back at you? This week’s Coffee Break AI Challenge is here to save the day! It shows you how to use AI to help you give clear and targeted feedback on pupils’ writing, all while staying true to the KS2 Teacher Assessment Framework (TAF).
The Challenge
Can we use AI to analyse a piece of student writing and generate feedback linked to the TAF for writing (working towards, expected, or greater depth)?
Step 1: Capture the Writing
For handwritten work you can use your phone’s camera to photograph the work. Apps like Files on iPad allow you to straighten and enhance your photograph to help the AI to recognise the text more accurately. Use the plus button to upload the image or, If it’s typed, copy and paste the text directly into ChatGPT or Google Gemini.
Step 2: Use the Right Prompt
Before you enter a prompt it’s always a great idea to tell the AI what role it is playing. You’ll see at the start of the prompt below we establish its role first. After the first student you don’t need to be so specific about the background to the request. You can also supply a link to the framework so that the AI can do it’s pre-research but I found in most instances it finds it anyway.
Prompt:
“You are an experienced primary teacher using the UK Teacher Assessment Framework (TAF) for writing. Analyse the following piece of student writing and provide feedback against the TAF statements for key stage 2. Make the feedback suitable for a Year 5 child.Please include:
- A brief summary of the child’s strengths.
- Next steps for improvement.
- Which TAF statements are securely met, partially met, or not yet evident.
- A short, student-friendly teacher’s feedback comment.
Step 3: Review and Personalise
- The AI will generate detailed, criteria-based feedback in seconds.
- You can ask it to simplify, shorten, or adapt the tone — for example:
“Make the feedback suitable for Year 5 language.”
“Turn this into three clear next steps for the pupil.”
Bonus Tip
You can even upload multiple samples to track progress over time. Try prompting:
“Compare this pupil’s writing to their previous piece. Has their control of tense, sentence variety, or punctuation improved according to the TAF?”
Time Taken
About 5 minutes per pupil — quicker than tracking down your favourite mug by looking in all the places you usually forget you’ve left it.
Why It’s Worth Trying
Next time you’ve got a marking mountain and five minutes to spare, Have a go and let AI lend a hand. Used well It’s like having a teaching assistant who knows the TAF inside out (and doesn’t drink all the coffee). Having an AI feedback partner can give you valuable inspiration to help support the formative assessment process and it really helps you articulate your teacher instincts faster. Remember though you are still the professional and your insight is paramount. The fact that you know the child makes you a vital guide on their learning journey. If the process of using AI to support giving feedback moves from providing inspiration to you to replacing you that might be a step too far.
Ethical considerations
Maintain your professional judgment as the final authority in providing feedback. It’s very useful to use AI as a tool to improve your efficiency not as an autonomous decision-maker. After all the last thing we want to do is prove to anyone we aren’t a vital component of the learning process and do ourselves out of a job 😉
There are ways you can avoid that happening:
- Review and Edit: Always review, adapt, and personalise the AI-generated feedback. Add your own context, clarifications, and specific examples that demonstrate you’ve read and understood the individual student’s work.
- Retain Control: View the AI output as your draft; your professional input ensures the feedback is accurate, appropriate, and aligned with exactly what your learner needs.
- Use it for efficiency: It’s great for tasks that genuinely reduce workload, such as drafting comments or aligning observations to standards.