The first wave of “AI in education” was mostly about catching students using ChatGPT to cheat. The more interesting story now is what happens when teachers and schools use these tools deliberately.
Personalized practice, not personalized curriculum
The strongest classroom results so far come from narrow, well-scoped tools — adaptive math and reading practice that adjusts difficulty question by question — rather than open-ended AI tutors. These tools free teachers to spend class time on discussion and problem-solving instead of drilling.
AI as a grading and feedback assistant
Teachers are increasingly using AI to draft first-pass feedback on essays — flagging structure, argument clarity, and grammar — which the teacher then reviews and personalizes rather than writing from scratch. Used this way, it’s proving to save meaningful grading time without removing the teacher from the loop on the final judgment.
The cheating problem hasn’t gone away — but detection is shifting
AI-detection tools remain unreliable enough that many schools have moved away from relying on them alone, instead redesigning assignments (in-class writing, oral defenses of take-home work) to make undisclosed AI use less useful as a shortcut.
What to watch for next
- AI teaching assistants for large lecture courses, answering routine questions so instructors focus on complex ones.
- School-level AI literacy curricula becoming standard, teaching students how these tools work and where they fail, not just how to use them.
- Clearer district policies on when AI use must be disclosed in student work, replacing the current patchwork of teacher-by-teacher rules.
Leave a Reply