OpenAI has quietly discontinued its AI-writing detector, citing a "low rate of accuracy." The AI classifier, launched in January, was shut down on July 20. It was designed to help users differentiate between human-written and AI-generated text. OpenAI says the tool correctly labeled 26% of AI-generated text and mislabeled 9% of human-written text as AI-authored. - The company plans to develop more effective methods for identifying AI-generated audio and visual content.
- AI-generated content detection has been a major concern in education, especially with tools like ChatGPT, which have raised worries about academic integrity.
- Researchers at the University of Kansas claim to have developed a method with over 99% accuracy to identify AI-generated scientific writing.
Meanwhile, a recent study showed that tools purporting to detect AI-generated text can be easily fooled. - The study evaluated 14 detection tools, including Turnitin, GPT Zero, and Compilatio.
- While the tools accurately detected human-written text 96% of the time, the accuracy rate of ChatGPT text was only 74% and 42% for slightly tweaked ChatGPT content.
|
