Detection Tools: Limitations and Alternatives

What You Should Know

Johns Hopkins is warning educators not to rely on AI detectors. Its teaching guidance points to questionable accuracy, false positives, and evidence that detectors can be biased against non-native English speakers. Instead, JHU recommends clearer AI policies, understanding students’ writing processes, and designing assignments that make learning visible.

Why should educators care? AI detection tries to answer the question after the fact: “Did this student use AI?” A better approach is to make AI use transparent from the start. Typerighter gives institutions visibility into the writing process and tools to set clear boundaries around AI use—reducing reliance on imperfect detection while giving instructors more information about how students actually developed their work.

The increase in availability of generative AI has brought forth a number of AI detection tools that claim to distinguish between AI and human-generated text. Turnitin, GPTZero, and Copyleaks are a few examples. Relying on predictable word patterns and sentence structure, these tools provide a percentage indicating how much of the text was allegedly produced by a human versus AI.

Understandably, these tools have generated a great deal of interest among instructors who are anxious to know how much their students are using AI to complete coursework. But how reliable are these tools? Despite lofty claims, the accuracy of these tools is questionable. [Read the full article.]