How to encourage smarter AI use in the classroom

Generative AI chatbots took schools by surprise, enabling students to answer homework questions or write essays instantly. Teachers often detect AI use through telltale mistakes or stylistic giveaways like excessive em dashes, but the technology has increased teachers' workloads as they juggle lesson planning, grading, and adapting to new tools. Despite encouragement from organizations like OpenAI and UNESCO, many educators remain unsure how to handle AI responsibly.

The article profiles Cheshire Academy, a private boarding school in Connecticut with about 400 students in grades 9–12. Administrators do not mandate AI use, yet librarian and technology coordinator George Aiello says the "vast majority" of instructors use it in some way. The school employs a patchwork of tools, including general-purpose chatbots like ChatGPT and Perplexity and specialized platforms like MagicSchool for educators. This approach stems from consultant advice to train staff on general AI techniques rather than prescribing specific technology. Training covered prompt-crafting skills and highlighted AI's limitations, including its potential for generating incorrect or biased responses.

Teachers at Cheshire Academy commonly use generative AI to prepare class materials, such as planning lessons or creating grading rubrics. Some have considered using AI to provide student feedback, but concerns about quality, personalization, and privacy have so far prevented them from doing so. French teacher Miriam Przybyla-Baum, who does not use AI herself, addresses it directly in her teaching. With nearly 30 years of materials, she noticed students relying on tools like Google Translate before ChatGPT's launch. In one assignment, she has students let a large language model edit their homework, then review the edits to decide which corrections are valid and which remove their voice.