Classic texts like The Great Divorce are exactly the kind of books that generative AI can summarize and analyze on demand, which puts teachers in a new position when assigning essays. Some students will use a chatbot to brainstorm, others to write whole paragraphs, and many will not know where the line is. A clear written policy removes the guesswork and keeps the conversation with students honest.

Begin by deciding what you actually want students to learn from the assignment. If the goal is for them to practice building an original interpretation of a scene, then outsourcing the thesis to a chatbot defeats the purpose. If the goal is to practice revising, then using a tool to spot unclear sentences might be perfectly reasonable.

Write the policy in plain language and include concrete examples of permitted and prohibited uses. A vague statement like "use AI responsibly" leaves too much to interpretation. Students respond better to specifics, such as whether they may ask a tool to explain vocabulary but not to draft a paragraph.

Spell Out What Is Allowed

Different assignments can call for different rules, so label each one clearly on the assignment sheet. For a close-reading essay on the Tragedian scene, you might allow AI for checking grammar but not for generating ideas. For a brainstorming exercise, you might allow broader use as long as students record what they asked and what they kept.

• Permitted: asking for definitions of unfamiliar words or historical references
• Permitted with disclosure: using a tool to suggest revisions to clarity or grammar
• Not permitted: generating thesis statements, body paragraphs, or conclusions
• Required: a short note describing any AI assistance used
• Always required: quotations and interpretations that come from the student's own reading

Design Assignments That Reward Original Thinking

Assignments that ask for personal engagement with specific scenes are harder to outsource. Requiring students to connect a passage to a class discussion, or to build on an annotation they made in their own copy, ties the essay to work only they did. Process artifacts like outlines and drafts also make it easier to see how thinking developed.

In-class writing, even a short one, gives you a baseline of each student's voice and skill. When a take-home essay differs sharply from that baseline, you have a reason to talk with the student, not to accuse. A conversation about how they developed their ideas is usually more revealing than any detector.

Grade the Work, Not Your Suspicion

AI detection software produces false positives and should never be the sole basis for a serious accusation. A fairer approach is to grade the essay against the rubric, and if you have concerns, ask the student to explain their argument and evidence out loud. Students who wrote the paper can usually do this easily.

Grading rubrics that emphasize specific textual evidence and original analysis naturally penalize generic, AI-flavored writing. An essay full of confident but vague statements about heaven and hell will score poorly because it lacks precise support. That outcome comes from the standards, not from an accusation.

Keep the Policy Alive

Technology and student habits change quickly, so revisit your policy at least once a semester. Ask students what they find confusing and adjust the language. A policy developed with student input tends to earn more respect.

Share the policy with parents and colleagues as well, so expectations are consistent across classes. When teachers in a department agree on core principles, students get a clear and steady message. That consistency reduces conflicts and lets you focus on teaching writing.