Every English teacher has read a paper that sounds impressive and says almost nothing about the book. The sentences are smooth, the vocabulary is elevated, and the claims could apply to any Gothic novel. With AI writing tools now common, that experience has become more frequent.

It helps to be honest about what you can and cannot know. Detection software is unreliable enough that many schools caution against using it as proof of misconduct. A more dependable approach is to design assignments and grading that reward what a student who actually read the book can do.

The same patterns that flag an AI draft often flag a paper built from study guides or online summaries. There is a generic theme statement, evidence that is described but never quoted, and no sign the writer has engaged with a specific passage. In all these cases, the underlying problem is a lack of close reading.

That framing keeps the conversation with students grounded in the work. Instead of accusing anyone, you can say that the essay does not yet show specific engagement with the text. That is a fair, rubric-based observation and it points to what needs to change.

Signals Worth Noticing

No single sign is proof, but a cluster is worth a closer look. A paper that gets small details wrong, refers to events out of order, or describes scenes the book does not contain suggests the writer never met the text. The reverse also happens: an essay that sounds nothing like the student's in-class writing deserves a friendly conversation.

• Broad claims about themes with no chapter, narrator, or document identified
• Quotations that are missing, invented, or slightly off from the text
• Plot details that do not match the book, such as the wrong character in a scene
• A voice and vocabulary very different from the student's earlier work
• Neat structure with no trace of a real struggle to figure something out

Design Assignments That Ask for Specifics

Prompts tied to particular documents, chapters, or narrators make it harder to rely on outside summaries. Ask students to analyze one journal entry, one letter, or the phonograph diary. Require page or chapter references so you can verify each quote.

Add process to the assignment as well. Collect a thesis, an annotated passage, and a rough draft before the final essay. Students who did the thinking can talk about it; students who did not will struggle.

Use In-Class Writing as a Baseline

A short handwritten or supervised response early in the unit gives you a sample of each student's voice and level. When a polished essay arrives that does not match, you have context for a conversation. It also lets you see who understands the book without relying on anything else.

Some teachers also ask students to explain a paragraph aloud in a two-minute conference. A student who wrote the paper can usually do this easily. It is a low-drama check that respects students while keeping the assignment honest.

Let the Rubric Do the Work

A rubric that rewards specific, accurate, well-explained evidence will naturally score generic essays lower. Tools like GraideMind that grade against those criteria can highlight where evidence is thin or unexplained, which supports a fair, consistent response. The teacher still decides whether something more serious is happening.

Keep the tone of feedback focused on the writing. Saying that an essay needs more text-specific analysis is accurate and constructive, and it gives the student a real path forward whether the weakness came from a shortcut or from simple uncertainty.