Island of the Blue Dolphins is one of the most widely written-about children's novels, which means AI chatbots know it well. A student who pastes the prompt into one can get back a tidy five paragraph essay in seconds. Teachers are noticing, and many are wondering how to respond.

Generic AI essays share a pattern. They stay general, describe Karana in broad terms like resilient and courageous, and avoid the small specifics that come from actually reading. There are rarely references to particular scenes, and the explanation tends to be polished but empty.

That gives teachers a way in. Prompts that require specific scenes, chapter references, or connections to class discussion are harder for a chatbot to fake. Asking students to connect the novel to something you read together, or to a moment that surprised them, raises the bar for authenticity.

Process also matters. Requiring a planning sheet, a handwritten draft, or a short in-class writing sample gives you material to compare against the final submission. If the voice changes dramatically, a conversation is warranted.

Design Assignments That Reward Specificity

Build the rubric so that generic writing scores lower. Add language such as "uses specific, accurate details from at least two distinct parts of the novel" and "explains each in the student's own words." A polished but vague essay then earns a middling score for the right reason.

• Require references to specific scenes or moments in the book
• Include a planning stage that is collected or checked
• Add a short in-class writing task to establish a voice baseline
• Ask for connections to class discussions or activities
• Score specificity and depth of explanation explicitly

Have the Conversation First

Be clear about your classroom policy on AI. Some teachers allow it for brainstorming and ban it for drafting, while others ban it entirely. Whatever you choose, students should hear it in plain language before the assignment.

If you suspect a paper is not the student's own work, start with a conversation about the content. Ask them to explain their argument or talk through a scene they cited. Their answers usually tell you what you need to know.

A Word on Detection Tools

AI detection software is unreliable and can flag honest students, so treat it as a weak signal at best. Grading for specificity and depth is more dependable. A rubric-based tool like GraideMind can score against those criteria consistently, which helps generic essays receive the scores they earn.

Keep the focus on learning. Students who see that thoughtful work is rewarded are more likely to do it themselves. A rubric that values real reading and real thinking does more than any policy memo.