Teachers have noticed that broad prompts about greed in literature are exactly the kind of questions generative AI answers most smoothly. A student can paste a generic question about Tvrdica into a chatbot and receive a plausible essay in seconds. Rather than racing to detect such work after the fact, many educators are redesigning assignments so that real engagement with the text and the class is required. The best defense is a task that only someone who has done the thinking can complete.

Specificity is the most effective tool. Instead of asking what the play says about greed, ask students to choose one scene and explain how a single exchange between Kir Janja and another character reveals his priorities. Require them to quote the specific lines in the translation used in class. A generic generated answer will struggle with this level of detail, particularly for a play that is not as widely documented as more famous works.

Connecting the prompt to classroom experience adds another layer. Ask students to respond to a point raised in discussion, to build on a peer's observation, or to apply an idea from a class activity. These references are unique to your classroom and cannot be fabricated by a tool that was not present. They also encourage attentive participation, which is a worthwhile goal in itself.

Process Based Assignments

Breaking the assignment into stages makes the thinking visible and harder to outsource. A sequence might include a brainstorm, a thesis checkpoint, a draft with peer feedback, and a short reflection on revision choices. Each stage is graded lightly, and together they show how the final essay developed. Sudden leaps in quality or voice between stages become easy to notice without any accusation.

• Require specific scenes and quotations from the assigned translation
• Tie the prompt to an idea raised in class discussion
• Collect a thesis and evidence plan before the full draft
• Include a short reflection explaining revision choices
• Add a brief oral or written follow up on the essay's main argument

Setting Clear Expectations About AI Use

Ambiguity about AI creates problems for everyone. Tell students exactly what is allowed, such as using a tool to check grammar or brainstorm questions, and what is not, such as generating paragraphs or analysis. Put the policy in writing on the assignment sheet and review it aloud. When students understand the line, honest ones can follow it and dishonest ones lose the excuse of confusion.

Explain the reasons behind the policy, not only the rules. Writing about a play is meant to develop the student's own ability to interpret and argue, which is lost when a tool does the thinking. Students tend to respond better to a rationale connected to their own growth than to a bare prohibition. This also opens the door to honest conversation about where technology can help.

Using Short Follow Up Checks

A brief conversation about the essay is one of the most reliable ways to confirm understanding. Ask a student to explain their thesis and walk through one body paragraph, which takes about three minutes. Students who wrote their own work can do this easily, while those who did not often struggle. Use these conversations sparingly and with care, treating them as a standard part of the process rather than an interrogation.

In class writing can serve a similar purpose. A short timed paragraph about one of the ideas in the take home essay provides a comparison point for voice and thinking. If the two pieces are wildly different, a conversation is appropriate. This approach relies on evidence and fairness rather than suspicion.

Grading With Integrity in Mind

Rubrics that reward specificity, original interpretation, and use of class material naturally favor authentic work. Generic, polished essays that lack concrete detail will score lower without any special detection step. This aligns grading with the learning goals and avoids over reliance on detection tools, which can be unreliable. Students quickly learn that real engagement earns better results.

Remember that detection software alone should never decide an academic integrity case. Combine multiple forms of evidence, including drafts, conversation, and comparison with prior work. Handle concerns privately and give students a fair chance to explain. A thoughtful process protects both the integrity of the course and the dignity of the student.