If you are a school or district administrator or literacy leader, you’ve probably noticed that the teachers who are often the most hesitant about incorporating AI into their curriculum are English teachers. We understand that hesitation. Both of us were English or ELA teachers at some point in our careers, and we know the magic that happens when you connect with a student around their writing. It’s a beautiful thing to convince a young writer that their life is worth writing about, their ideas matter, their voice is significant. It’s fulfilling to coach that writer through stages of revision so that they emerge not only with a stronger piece, but as a stronger writer.
On the other hand, we also know what it’s like to be a teacher with 60–100 students, wanting to give intense, intimate, insightful feedback to every writer but constantly bumping into roadblocks like short class periods and too little time. We know that frustration when you realize that these roadblocks mean that most of your writers don’t get quality feedback early in the writing process.
AI can help. It can help students and writing teachers. We are both teachers and professional writers, as well as AI researchers, and we say with confidence that one of the things that AI is best at is working with writers. We hope to instill that same confidence in school leaders, so they can strategize with teachers about AI and create conditions in schools where all students flourish as writers.
The Virtual Writing Coach
We know from John Hattie’s research (2023) and our own classroom experiences how important it is for feedback to happen swiftly. Yet time constraints often make it difficult for teachers to meet with all writers early in their drafts. This is where AI comes in. If your school invests in high-quality, data-protected AI tools such as Flint, MagicSchool AI, or Gemini, your writers can be invited into a coaching conversation with an AI partner as they write. You can also set boundaries with the tools—such as “Do not draft for students,” “Do not find evidence for students,” or “Ask students to share their thinking and work first”—that can train students in brain-first habits. These practices prioritize students’ own thinking, analysis, and problem solving before they seek external support from AI or other tools (Kosmyna, 2025).
For an example of what these coaching conversations can look like, let’s consider a draft by Lauren, a 5th grade journalism student who is writing about her own experience talking to peers about her neurodivergence. She writes her draft as a news event and takes up a journalist’s tone.
Your writers can be invited into a coaching conversation with an AI partner as they write.
As you read, ask yourself what kind of feedback you would hope that Lauren might receive and what kind of tone that feedback should take. Supportive of her vulnerability while also pushing her craft forward? Honoring the risk she’s taking while helping her develop as a journalist?
The meeting began at 9:00 a.m. sharp. The room was quiet as Lauren began to speak. She explained that she can’t remember things because her brain was injured. She also explained how it was getting better because she was working on rewiring parts of her brain. Lauren showed pictures of brain activity. She spoke about the medications she took. She described what it was like when she suddenly wouldn’t remember where she was, like if it was an unfamiliar place—like when she first came to 5th grade.Everyone in class wanted to know how her brain got injured. “When I was a baby,” she explained. A boy asked if it could happen to anyone. Lauren explained that brain injuries can happen if you get hit on the head or if you don’t have enough oxygen. “Then parts of your brain don’t work and you have to train other parts.” Kids were quiet. “I hit my head in football,” Brian said. “I had a concussion last year,” said Sean.This reporter remembers that Sean often teased Lauren for being slow.Brains can get hurt. So can people.
When we read this piece, we wanted Lauren to get warm, affirming feedback. Here’s how her AI partner (in this case, Flint), responded to this early draft. The chat was designed to invite students to share their early writing and to affirm their strengths. The teacher directed AI, “Notice what is beautiful in this writing and compliment the writer.” Also, the teacher clarified with the AI tool, “Your role is to help this writer see their own power.”
AI’s response was as follows:
What a powerful and compassionate piece of journalism! Here are three things that really stand out:1. Your storytelling brings us right into that classroom moment. The way you capture the quiet room by writing as a journalist, the brave explanation that you, Lauren, give, and the specific details like the brain activity pictures makes readers feel like they’re sitting there with you. You have a real gift for painting a scene with words, especially noting important moments like “The meeting began at 9:00 a.m. sharp.”Your courage in highlighting both kinds of hurt is remarkable. By remembering and including that Sean had teased before, then showing his own vulnerability about his concussion, you’re helpingeveryone understand that we all have struggles and need kindness. That takes real bravery as a reporter and a writer to include that honest observation about past unkindness.2. Your closing line is powerful. “Brains can get hurt. So can people.” This shows incredible wisdom and empathy beyond your years. You’ve connected physical injuries with emotional ones in just seven words, advocating for everyone who faces challenges. This is the kind of powerful writing that changes how people think and treat each other.3. You’re not just developing as a strong writer—you’re using your voice to build understanding and compassion in your classroom. That’s what the best journalists do.
The first thing to consider as a literacy leader is: Are you happy with this initial feedback? We are. The AI partner gave feedback that affirmed the student’s courage and her power as a writer and a young person advocating for dignity and kindness. Lauren’s AI partner, working with context provided by her teacher, recognized that she is in her early stages as a writer and asked if she would like some possible next steps. When she agreed, AI suggested:
Here are three next steps that might help you grow this piece and grow as a writer and journalist.1. Interview and add more voices. Consider talking to classmates or a teacher about what they learned from the presentation. You might respectfully ask Sean how his perspective changed. Adding these voices would make your journalism even richer and show multiple viewpoints—a key skill for reporters.2. Expand the “after” story. Your piece captures the meeting beautifully, but readers might wonder: What happened next? Did kids treat Lauren or others differently? Did Brian or Sean share more about their own experiences? Adding a paragraph about the days or weeks following this presentation could show the real impact of your courage in speaking up.3. Create a “reporter’s notebook” for yourself. You have excellent instincts for noticing important details (like the quiet room, specific times, and remembered interactions). Start keeping a small notebook where you jot down observations throughout your day—interesting phrases people use, moments of kindness or conflict, details that catch your eye. This practice will make your already-strong observational skills even sharper and give you a treasure chest of details for future writing.Your compassionate eye and brave voice are exactly what journalism needs. Keep noting not just what happens, but what it means—you’re already doing this beautifully with your connection between physical and emotional hurt.
Such feedback can not only teach students how to improve their piece, but also increase their confidence as writers, thinkers, and young activists (Ehrenworth & Seyfried, 2025). When we’ve seen students interact with AI in this way, they look engaged, often pointing out favorite parts of the digital conversation to their peers and jotting notes of what they want to take away.
And because students who need affirmation the most often get it the least (Howard et al., 2020), an AI partner can act as an assistant writing coach, ensuring that all students get some immediate, warm, specific feedback. Then the teacher’s role becomes less about dishing out the feedback and more about reviewing and expanding upon it with students, such as Lauren’s teacher saying, “I saw the feedback your AI partner gave you, and I so agree. You show real courage as a journalist. Let me teach you something about interviewing—you might practice with your partner, Sarah, first.”
An AI coach can also respond to students in virtually every language and can translate their work so teachers can see multilingual students’ full competencies. Further, these tools can support teachers in communicating about student work to families and caregivers who might speak a different language at home.
Helping Teachers Give Better Feedback
The AI response given to Lauren earlier in this article demonstrates a few qualities of effective feedback that we have often taught in our graduate classes. Good feedback:
- Begins by summarizing and elevating what the student has done, quoting specific lines and phrases to highlight the student’s words.
- Describes the writing work that the student has done—the powerful moves she has made, so she can strive to replicate them in other pieces.
- Honors the student’s perspective, humanity, and identity.
- Grounds next steps and tips in what the writer has already started, so the feedback is asset-based.
- Supports the writer in not only expanding and improving this piece of writing, but also in growing as a writer and thinker overall.
With this type of effective AI feedback as a starting point, teachers can use technology as a teaching assistant and then go deeper with their own feedback to students. They can ask themselves:
- What’s beautiful about this writing that I might share with this young writer to build their confidence?
- What writing moves has this student made that show promise, even if the piece isn’t polished yet?
- If this writer keeps developing the strengths visible in this draft, what kind of writer might they become?
Schools might also invest in technology such as Flint or MagicSchool AI and give teachers time and training on how to create effective learning opportunities within these chats. (See Figure 1 for tips on how to set these chats up.) The important thing is that students have access to quality feedback at all stages of the writing process and get affirmation about their strengths and areas to improve, both with a responsive technology tool and their human teacher.
Two Essential Leadership Questions
As leaders consider how AI might support writing instruction across their buildings, there are essential questions worth exploring with the leadership team, teachers, and school community. These aren’t questions with single right answers, but rather invitations to think together about implementation.
1. How can we get all teachers involved in leveraging AI to make writing more accessible and feedback more universal?
Not every teacher will feel equally ready or willing to explore AI as a teaching partner. Some will be early adopters, excited to experiment. Others will need time, support, and evidence before they’re comfortable. Consider starting with interested volunteers who can pilot AI writing partnerships in their classrooms, document what they learn, and share their experiences with colleagues. Professional development might include teachers trying AI themselves as writers first and experiencing what it’s like to get feedback from an AI partner, noticing what feels helpful and what doesn’t. When teachers experience AI as learners, they develop stronger instincts about how to guide students in using it.
AI can provide initial, affirming, targeted feedback, freeing teachers to spend more time working with students to make sense of feedback and take action on it.
You might also create space for teachers to design AI chats together, thinking through what brain-first principles look like in their specific grade levels and content areas. Perhaps most important, build cultures of confidence and trust by creating AI sandboxes where students and teachers can experiment in low-stakes environments. When schools focus on surveillance and mistake-prevention, they create what Eduardo Briceño (2023) calls “performance zones”—environments where fear of judgment shuts down the very risk-taking that leads to learning. In contrast, “learning zones,” where experimentation is safe, foster the intellectual courage students and teachers need to grow.
2. What if we could shift some of teachers’ time from producing feedback to helping students implement it?
This is one of AI’s most compelling possibilities. Teachers spend countless hours writing feedback that students often struggle to apply. AI can provide initial, affirming, targeted feedback, freeing teachers to spend more time working with students to make sense of feedback and take action on it. This is where real teaching happens—in coaching conversations where educators help student writers see what revisions can make their writing stronger. AI can’t replace that human connection, but it might make more of those moments possible for every student.
For example, Lauren’s teacher might say, “AI noticed your closing line is powerful. Let’s look at that together—what were you trying to do there? How might you build toward that moment earlier in your piece?” Or with another student, “AI said your evidence is strong. Show me where you see that strength. Now, where else in your draft could you use evidence like that?”
AI can name what a student is doing well and what they are ready to work on next, but it is the teacher-student discussions and interactions surrounding the writing that makes it meaningful.
Encouragement for Emerging Writers
We began by acknowledging the hesitation many English teachers feel about AI. The work of connecting with a young writer, of convincing them their voice matters, of coaching them through revision is deeply human work. AI doesn’t replace that. But it can help make that work possible for more students, more often. In classrooms where AI provides initial feedback, we’ve seen students who previously gave up on revision start asking questions like, “How could I make this part stronger?” We’ve watched students who were previously labeled “reluctant writers” lean into their drafts because someone—even an AI someone—noticed what they were doing well. We’ve seen students like Lauren, navigating deeply personal topics, receive the encouragement they needed to keep writing through hard things. Our students, all of them, deserve that kind of attention to their ideas, their voices, and their growth as writers and thinkers.
Reflect & Discuss
Do you agree with the qualities of good feedback that the authors outline? Why or why not? What would you add to that list?
What benefits do you see for allowing AI to offer initial feedback on student writing? What drawbacks?
