The AI detectors got it wrong more than half the time when evaluating writing from non-Native English-speaking students, recent research found.