Peer review has long been a staple of writing instruction, but its quality varies enormously. Students often give vague comments like "it was good" or fixate on spelling while missing larger problems in an essay on "Second Variety." AI-generated feedback can supply a more structured perspective, but it lacks the sense of a real reader and the social accountability of a classmate. Used together, they can cover each other's weaknesses.

A layered workflow might begin with the student receiving AI feedback on a draft aligned to your rubric. The student then reviews that feedback, decides which suggestions make sense, and revises. Only after this first revision does the paper go to a peer, who can respond to the ideas rather than to basic problems. This sequence uses each source at the point where it adds the most value.

The teacher's role is to design the process and monitor it. You can check a sample of AI comments for accuracy, review peer comments for quality, and step in where students are stuck. This keeps the technology in a supporting role. It also protects students from relying too heavily on any single source.

Training Students to Give Better Peer Feedback

Peer reviewers need explicit guidance, or their comments will stay shallow. Give them a short checklist tied to the rubric, with questions such as whether the thesis is clear and whether each quotation is explained. Model a good peer comment and a weak one so they see the difference. A little training goes a long way.

• Identify the thesis and say whether it is clear and arguable
• Point out one place where evidence is strong and one where it needs explanation
• Ask one question that would help the writer develop an idea
• Avoid commenting on grammar until ideas have been reviewed
• Explain every suggestion so the writer understands the reasoning

Teaching Students to Evaluate AI Comments

AI feedback is a tool, not an authority, and students should learn to evaluate it critically. Ask them to mark which comments they agree with, which they disagree with, and why. This turns feedback into a thinking exercise and prevents passive acceptance. It also builds the kind of judgment students will need as AI becomes more common.

Make the expectations clear about authorship. Students should understand that feedback may suggest improvements but that the revision must be their own work. Setting this boundary early reduces confusion about academic integrity. It also keeps the focus on learning rather than on producing a flawless draft.

Measuring Whether It Works

To see whether this layered approach helps, compare drafts and final essays. Look for changes in thesis quality, use of evidence, and explanation of ideas. Ask students which type of feedback was most useful and why. Their answers can help you adjust the process.

Keep the system simple enough to sustain. If it demands too much time or technical effort, it will not last beyond the first unit. Start with one assignment, refine the steps, and expand gradually. A manageable process that students understand is more valuable than an elaborate one that confuses them.