Use AI to write personalized feedback by giving it three inputs: the student’s work, your evaluation criteria, and one or two sentences about what you noticed. Claude or ChatGPT then drafts detailed, specific feedback in your tone that you review and refine before sending.
The Feedback Drafting Method
The key to using AI for student feedback is giving it enough context to be specific rather than generic. A prompt like “Write feedback on this student submission” produces bland output. A prompt like “Write encouraging but honest feedback on this student’s first course module draft. They’re new to online teaching and nervous about going live. Criteria: clarity, practical examples, actionable advice. Here’s their draft: [paste]” produces something genuinely useful.
Your job as the educator is to supply the observation layer — what you actually noticed about the work — and let AI handle the drafting layer. You might write two sentences of raw notes: “Good structure but the examples felt generic. Needs more specificity about their actual audience.” AI expands those notes into a full paragraph of warm, constructive feedback. You review, edit, add your voice, send.
Scaling Feedback Without Losing Quality
In a cohort of 20 students, writing meaningful feedback on every submission used to take hours. With AI drafting, the same task takes 30–40 minutes — about 2 minutes per student. That’s enough time to read the submission, jot two sentences of observations, generate the draft, edit once, and send. The feedback is still personal because your observations drive the content; it’s just no longer limited by how fast you can type.
You can also build reusable feedback templates for common patterns — the student who lists everything but doesn’t apply it, the student who second-guesses their expertise, the student whose structure is strong but writing is dense. Store these in Claude’s projects or your notes app and adapt them with specific details for each student. The template saves you from starting from scratch; the details make it feel personal.
What This Means for Educators
Feedback quality directly affects student outcomes. Students who receive specific, encouraging feedback apply more, improve faster, and stay enrolled longer. If AI lets you give better feedback to more students in less time, that’s a direct improvement to your program’s impact — not just an efficiency win. Use AI to raise your feedback floor, not to automate it into meaninglessness.
The Simple Rule
Always write at least two specific observations about the student’s actual work before generating a feedback draft. Those observations are what make the feedback real. AI handles everything else — length, tone, structure, encouragement. You supply the eyes; AI supplies the words.