A short, well known text like Claude Gueux is exactly the kind of work that AI tools can summarize and analyze in seconds. Students who have not read the story can generate a plausible essay with generic observations about poverty and injustice. Teachers cannot rely on detection software alone, so the more durable solution is to design assignments that require genuine engagement with the text.

Prompts that require specific, verifiable details are harder to fake. Ask students to analyze a particular exchange between Claude and the director, cite where it appears, and explain how a single word choice in your assigned edition shapes the meaning. Generic summaries will not satisfy a task grounded in a specific passage and edition.

Incorporate personal and classroom specific elements that an outside tool cannot know. Reference a discussion from class, a passage you annotated together, or a question a classmate raised during seminar. Students must then connect their analysis to the shared experience, which makes the work more distinctive and more meaningful.

Building Process Into the Assignment

Requiring visible stages of work makes it harder to submit a finished essay produced elsewhere. Ask for annotated passages, a short proposal, a handwritten or in class outline, and a draft with a revision memo. Evaluating these stages also gives you a clear picture of the student's thinking.

• Require annotations or notes on specific passages before drafting begins
• Hold a brief in class writing session for the thesis and opening paragraph
• Ask for a revision memo explaining what changed between drafts
• Include a short oral check in where students explain their argument
• Tie the prompt to class discussions and shared materials

Setting Clear Expectations

Students are more likely to use AI inappropriately when the rules are unclear. State in writing which uses are acceptable, such as asking a tool to explain a difficult word, and which are not, such as generating paragraphs for submission. Explain the reasoning so students understand that the policy protects the value of their own learning.

Discuss the limits of AI generated analysis openly. Show a sample AI essay on Claude Gueux and ask the class to critique it, noting where it stays vague or misses details. Students often find this exercise eye opening, since they see how generic the writing can be.

Grading With Integrity in Mind

Rubrics can reward the qualities that AI generated work often lacks, including precise evidence, original interpretation, and engagement with the specific class context. Include criteria for specificity and for connection to classroom discussions. Essays that rely on general statements will naturally score lower without requiring accusations.

When you suspect a problem, begin with a conversation instead of a penalty. Ask the student to explain their argument and point to the passages they discuss. A student who wrote the essay can usually do this easily, and the conversation is more productive than relying on detection scores.

Using AI Constructively

Rather than banning AI entirely, consider structured uses that support learning. Tools like GraideMind give students rubric aligned feedback on their own drafts, which encourages revision without producing the writing for them. The distinction between feedback on student work and generation of student work is an important one to teach.

Revisit your policies each term as tools and student habits change. Gather feedback from students about how they use AI and what they find helpful or confusing. A flexible, transparent approach keeps your assignments relevant and your expectations fair.