Imagine it’s 2:00 a.m., and you can’t sleep because you’re thinking about all the essays you need to give feedback on. As a secondary English language arts teacher, you might have up to 150 pieces of student writing to respond to. You want to give all your students multiple rounds of timely, specific feedback. But there just aren’t enough hours in the day, and you can’t keep up.
I’ve been that teacher, and I know.
Researchers have described this work of providing feedback, particularly formative feedback, as an “unmanageable task” (Sloan & Peters, 2022). Let’s say that giving feedback on an essay takes 10 minutes; if the teacher has 100 students, it can take more than 16 hours to complete one round of drafts. When you factor in all the other demands on classroom teachers, it’s no wonder that many describe their workload as “unbearable,” with some leaving the profession altogether (Elevate K–12, 2024).
But take heart, because recent advancements in artificial intelligence (AI) can help. Students can use AI to get specific, timely feedback that they can immediately act on—and this can happen without sacrificing instructional quality, teacher judgment, or student ownership. Educator Nathan Lang-Raad (2024) suggests using AI to “provide a level of individualized attention and feedback that would be impossible for a teacher to provide alone,” including “feedback tailored to [students’] specific writing challenges.” This perspective positions AI as a powerful tool to increase teacher capacity. Students can use AI to write more, revise more purposefully, and engage more closely with the writing process. At the same time, AI can not only help teachers create a more manageable workload, but also enable them to engage more deeply with students about their work.
Moving from Distrust to Development
When AI burst into the mainstream in late 2022, the possibilities for learning seemed endless. However, given students’ early use (and abuse) of the tool, educators soon reacted with fear and tried to find ways to prevent students from turning to AI to do their work for them. As teachers, we wanted our students to convey their thoughts and emotions, make meaning, and find their voice through the written word. The idea that they could simply ask AI for an essay that answered our prompt—and then copy and paste it into a document to turn in—not only was disheartening in terms of students’ learning, but also felt like a threat to our identity as teachers of English.
So, what did teachers do to combat this threat? Many switched to using quick, in-class assignments that they could monitor for the use of AI. They used plagiarism checkers, such as Turnitin, and rewrote classroom plagiarism practices to clarify that if students used AI in their essays, they would earn a zero.
The truth is, students have been using AI tools for years. When a Google document suggests the next word or phrase in a sentence, it’s using AI. When students rely on Grammarly to check their syntax and word choice before turning in a finished draft, they’re using AI. ChatGPT and other artificial intelligence tools are simply the next iteration in programs that our students will learn to use over their lifetime.
More recently, our focus as educators has shifted from asking, “How do we stop students from using AI?” to “How can we help students use AI in ethical and effective ways?” There’s also another important question for us to consider: “How can AI make us teachers more effective while making our jobs more sustainable?”
To write an effective prompt, students will need to identify who they are as a writer, what type of writing they’re doing, and what specific feedback they’re looking for.
It’s About Support, Not Replacement
The traditional writing classroom focuses on students producing work, peers and the teacher giving initial feedback, and then the student revising and submitting a finished product for summative feedback from the teacher. We can reframe the writing process to add a layer of support from AI to the feedback loop. To do this successfully, we must teach students how to prompt AI to give meaningful feedback that aligns with our expectations.
The most effective student partnership with AI will start with the student writing a complete first draft on their own. Studies have shown that this step—thinking independently and writing first—is key to student learning. A study led by the Massachusetts Institute of Technology (Thompson, 2025) found that students who write without AI assistance but then revise with AI have “the strongest brain-wide connectivity,” which suggests increased learning is taking place. Alternately, when students start writing with AI and then work independently, they not only don’t show the same brain activity, but also report “lower satisfaction and ownership over their work.” Therefore, it’s essential to help students understand why they must write first, before looking to AI.
In truth, even our best and most thoughtful efforts to teach students the value of the writing process and of their own words may not reach everyone. A few students will probably continue to use AI as a replacement rather than an assistant. To allow for easier monitoring, we could use this as an excuse to have students do all their writing in class. But we shouldn’t fall victim to this line of reasoning. If we truly want to convey our belief in student thinking and the written word, we must value the writing process. We must give students the time and space to write when and where they wish.
Once they’ve written their first draft, each student should identify a few areas they want feedback on. They should consider parts of the essay they’re unsure of, such as whether the transitions between body paragraphs are effective or whether the thesis directly answers the prompt. They might also consider skills that some of their previous writing efforts have shown they need to hone. For instance, if the student received previous feedback about quote integration, they might decide to ask for feedback on whether the sentence structure of their lead-ins flows seamlessly into the quote or whether they have cited the quotes correctly using Modern Language Association (MLA) format. Alternatively, they might zoom in on skills that the teacher is highlighting and assessing during a given unit, such as identifying how a complex character changes over the course of a text. The student might want to ask for feedback on whether their chosen quotes and their analysis of those quotes fully explain how the character has changed. Specifying the feedback they need gives students a sense of ownership over the revisions and their work as a whole, as opposed to letting an AI tool decide for them what’s important.
Writing the Prompts
With those areas identified, students are now ready to ask AI for feedback using clear and specific prompts. In the beginning, they’ll need instruction about how to write a good prompt and will benefit from looking at sample prompts. The teacher may even provide sentence stems or a “fill in the blank” prompt for students to use. Here’s one sample prompt:
I’m a 9th grade English language arts student who wrote a literary analysis essay about Fahrenheit 451. Review my essay and provide the following feedback:1. Tell me two specific areas where my analysis of quotes could be stronger. Explain why these areas need improvement and offer brief, actionable suggestions.2. Tell me two specific places where I could add or improve transitions between ideas or paragraphs.Do not rewrite the essay; only explain what needs revision and why.
Here’s another prompt:
I’m a high school student who has written a personal narrative. Review my essay and provide the following feedback:1. Tell me two areas where my storytelling would improve with more vivid language, description, dialogue, or reflection. Explain why these areas need to be better developed.2. Tell me two places where my pacing or organization could be improved. Explain what I could change to clarify the sequence of events.Do not rewrite the essay; only identify the areas of need and explain what kind of revision would improve the narrative.
To write an effective prompt, students will need to identify who they are as a writer, what type of writing they’re doing, and what specific feedback they’re looking for. For example, in my writing class, a student was writing a feature story about teachers at our school who have spent their entire careers in the building. When I reviewed the student’s draft, we agreed they needed to work on providing evidence and explaining their reasoning, as well as on style and mechanics. The student crafted the following prompt:
I’m a high school journalism student who has written a feature. Review my essay and provide the following feedback:1. Tell me two areas where I could improve style or mechanics. Explain why these areas need to be revised.2. Tell me two ways I could improve my argument.Do not rewrite the essay; only identify the areas of need and explain what kind of revision would improve the piece.
ChatGPT provided the following feedback about style and mechanics:
Shifts between past and present tense. Example: “From the time when it was built, some of the school’s halls and rooms have never been modified.”Why this needs revision:There are moments where the narrative moves between past and present in ways that can create mechanical inconsistency. In the example cited, the phrasing combines past reference (“from the time when it was built”) with present-perfect (“have never been modified”), which can sound awkward or imbalanced.What type of revision would help:Ensuring consistent verb tense within each section—or choosing stronger transitions that make the tense shifts intentional—would smooth the mechanics and make the story easier to follow.
AI did what I would have done as a teacher; it provided a suggestion for improvement, reviewed why it was important, and then explained what kind of revision would address the issue.
ChatGPT also provided broad feedback about the argument, including this clarification: “Even though a feature story isn’t an argument-driven genre in the same way as an editorial is, your piece still makes claims about what has changed and what has stayed the same at LHS.” For example, the student had included teacher quotes in the article that were meant to show how the teachers’ experiences at the school shaped their present work there. ChatGPT provided the following feedback in this area:
Why this needs revision:The narrative would be stronger if you showed more how their experiences shape their work or relationships now. Right now, some quotes support the claim indirectly, but the argument is not fully illustrated.What type of revision would help:Adding or highlighting moments where the teachers describe a concrete example of how their past influences their present experience would make the argument clearer and more convincing.
This student was able to make purposeful revisions to her writing in ways that she wouldn’t have done on her own and without having to wait until I had the time to focus on her draft.
AI is one source of feedback—not the only definitive source of feedback. The classroom community is still a vital part of the process.
Bringing AI into the process is a real time saver for teachers. The teacher can hold quick writing conferences with each student without needing to take hours beforehand to read each essay in its entirety. The writing conference can focus on the areas the student wanted feedback on, how AI responded, and what choices the student was intending to make and why. The teacher can then encourage any further revisions or answer specific questions. This is particularly helpful if students aren’t entirely clear on what the feedback means or if they’re unsure whether to act on it. AI is one source of feedback—not the only definitive source of feedback. The classroom community is still a vital part of the process.
As students learn how to incorporate AI into their writing practices, some will, of course, use AI in ways that provide more assistance than the teacher intended. A student might ask AI to rewrite a section of the essay or choose better quotes instead of asking for feedback about how they might revise these areas. To better account for the learning process when AI is involved, have students submit their AI chat transcripts with their final essays; this leaves space for reflection about the writing product itself, about the writing process, and about the use of AI. Students and teachers can reflect together about what feedback was useful, why the students made certain decisions as writers, and what they might improve the next time around. By framing this as part of the learning and writing process, teachers can turn potential misuses of the tool into opportunities for students to learn another key literacy skill: ethical AI collaboration.
The Power of the Partnership
Partnering with AI during the writing process is a powerful practice for both students and teachers. Students can get specific feedback on their writing whenever they’re ready, without having to wait days for a teacher to read their draft. At the same time, because teachers are freed from having to read and comment on entire drafts, they have more time for deeper conversations with students about their ideas and use of various techniques.
It’s time to think of artificial intelligence as a useful tool in the writing toolbox.
Reflect & Discuss
As AI enters the feedback loop, which parts of teaching writing become more essential—and what risks being diminished?
What specific writing assignments or stages of the writing process would be most appropriate to start integrating AI feedback?
In what ways could requiring students to submit their AI chat transcripts alongside final essays serve as both an accountability measure and a teaching opportunity?
