How One School Got AI Right in the Classroom

When ChatGPT landed in classrooms, teachers panicked. Students could suddenly answer any question in seconds. But at Cheshire Academy, a small private school in Connecticut, they decided not to ban it. Instead, they trained teachers on how to use AI well — showing its strengths and its dangers, like making up facts or repeating bias. Teachers now use AI to plan lessons and create grading rubrics, saving hours of prep time.

The school also teaches students to think critically about AI. In one French class, students let a chatbot edit their homework, then decide which edits were good and which removed their own voice. In another, students grade each other's work while flagging parts they suspect were AI-written. The idea is to make students reflect, not just comply.

Assignments now carry traffic-light labels: green means AI is fully allowed, red means no AI at all, and yellow means some tools are okay (like spell-check) but chatbots are not. The school is also piloting a “Student AI Council,” where kids lead discussions about healthy AI use. It's a messy, evolving experiment — but it points to a more hopeful path than fear or blanket bans.