The Number That Finally Explains Why AI Detectors Keep Getting It Wrong
New research puts AI-detector false positives at 61% for non-native English speakers, too high for any score to stand alone as punitive evidence.
95,000 Students, One Study, and the Number That Should Change How You Talk About "AI Cheating"
A study of more than 95,000 students, published in Science, found that AI cheating clusters by discipline and by how often a student uses AI, not by whether a course allows it. Daily users cheat at 26%; monthly users cheat at 7%. What that means for anyone still writing a one-size-fits-all AI policy.
Why Detection-First Was Never Going to Work
A Stanford study found AI detectors misfire on nearly two-thirds of essays by non-native English speakers. Here is what to do instead of playing detective.