Grading and feedback are at the heart of teaching. They shape how students learn, reflect, and grow. But they also consume endless hours—hours that could be spent planning, connecting, or simply catching your breath after a long day in the classroom.

A new study suggests that AI might finally tip the balance in your favor. Researchers compared teacher-written feedback with feedback drafted by a Large Language Model (LLM) and lightly reviewed by teachers. The surprise? Students rated both types equally useful—and most couldn’t tell the difference.

---

The Teacher’s Weekend Dilemma

Picture this: It’s Friday evening. Ana, a high school chemistry teacher, has just carried home a stack of lab reports. She knows the feedback matters—her students need guidance, encouragement, and corrections to really learn.

But after hours of careful commenting, her notes get shorter and sharper. Some answers receive the thoughtful feedback they deserve; others get a rushed “Check your definitions” or “Review this again.”

Ana feels guilty. She wants every student to feel seen, but she’s out of time.

Sound familiar?

---

What the Research Found

The study, conducted in Brazilian secondary schools, tested feedback in chemistry classes. Sometimes students received comments written fully by their teacher. Other times, the teacher used an LLM-powered tool that suggested feedback, which they could edit or approve.

The results were striking:

• Students rated the AI-assisted feedback just as clear and motivating as teacher-only feedback.
• 85% of students couldn’t tell the difference—they assumed the AI-written comments were from their teacher.
• The AI-generated feedback was 2.6× longer than teacher-written feedback, yet teachers didn’t spend extra time reviewing it.

In other words, teachers were able to offer richer, more detailed guidance without increasing their workload.

---

Why This Matters for Teachers

For Ana—and for countless teachers like her—this is a game changer. Instead of writing every sentence from scratch, she could review and refine AI-suggested comments. Her role shifts from “scribe” to coach and curator of feedback.

The AI drafts the message. Ana ensures it fits her student’s needs, corrects the nuance, and adds her human touch. Students still feel her presence in every comment—because she’s still the final voice.

And because the AI saves time on wording, Ana can give more meaningful, specific guidance to more students.

---

What It Means for You

The takeaway from this research is simple:

• Your students value feedback, whether it’s AI-assisted or fully written by you.
• They often can’t tell the difference—as long as your voice guides the process.
• You can deliver longer, more personalized comments in the same amount of time.

At Assessain, we’re building tools to make this possible for everyday classroom work. You upload your worksheets, scan student answers, and our AI drafts feedback tailored to your rubric. You’re still in charge—but you don’t have to start from a blank page every time.

---

Try It Yourself

If you’ve ever wished your feedback could be richer, faster, and more consistent—without costing you your evenings—AI might just be the partner you’ve been waiting for.

👉 Start with Assessain today and see how it feels to have an assistant who drafts feedback so you can focus on what matters most: guiding your students’ learning.

---

📚 Citation:
Xavier, C., Torrezão da Costa, N., Valdo, A.K., Alves, G., Rodrigues, L., Rodrigues Neto, M., Silva, M., Falcão, T.P., Gašević, D., & Mello, R.F. (2026). Human Teacher vs. LLM-Generated Feedback in Secondary Education: A Comparative Study on Student Perceptions. In: Tammets, K. et al. (eds) Technology Enhanced Learning. EC-TEL 2025. Springer, Cham. https://doi.org/10.1007/978-3-032-03870-8_36