Many institutions acknowledge detector limitations while continuing to use them. That tension creates confusion for both students and faculty. Is reliance driven by necessity rather than confidence?
Academic writing requires paraphrasing sources and integrating ideas from multiple texts. Every time i paraphrase a complex source carefully, my ai score climbs. The thing detectors seem most suspicious of is exactly the intellectual work academics are supposed to be doing.
Ran a popular children's picture book through a detector as an experiment. Simple vocabulary, short repetitive sentences, clear structure. Flagged as likely ai. Children's writing is intentionally formulaic because that's how early readers process text. The detector has no concept of audience or purpose.