Lesson 108 · 3 min
Teachers: AI homework and fair judgment
AI detectors are unreliable, so fair teachers rely on conversation, process and well-designed tasks instead of a detector score.
AI text detectors give a score that looks precise but often is not. They miss AI text that was lightly edited, and they wrongly flag human writing. Studies have found they flag writing by non-native English speakers more often, which makes them unfair as evidence.
Treat a suspicion as a reason to talk. Ask the student to explain their argument, define a word they used, or describe how they wrote it. Someone who did the work can usually talk about it.
Design tasks that show process: drafts, notes, an in-class part, a short reflection on how they worked. Connect tasks to class discussions or local examples AI would not know.
Set clear rules about when AI is allowed and how to say they used it. Many students use it because the rules are unclear, and an honest policy invites honest answers.
Do it now
Write a clear AI policy
- 1. Think of one assignment you set regularly.
- 2. Use the prompt below to draft a short AI policy for it.
- 3. Share it with students before the next assignment.
Help me write a short, fair AI policy for this assignment: [DESCRIBE IT], grade [GRADE]. Say which AI uses are allowed (e.g. brainstorming, grammar feedback), which are not, and how students should say how they used AI. Use simple language, under 150 words, and add 3 questions I can ask a student to check they understand their own work.
Quick check
Keep going
Next →Turn a video into a lessonWords from this lesson
- Classifier
- A model that sorts things into groups, such as spam or not spam, or which plant is in a photo.
- Bias
- When an AI treats some people or ideas unfairly because of skewed patterns in the data it learned from.
- Watermarking
- Hiding an invisible signal in AI-made text, images or audio so software can later detect where it came from.