The emergence of AI-generated text has spurred a wave of detectors designed to spot content written by AI models . But can these applications really consistently tell the distinction between human and AI-generated writing ? Current analyses suggest a complicated reality: while some ai and human instruments demonstrate a measure of success, they are