A recent study conducted by the University of Kansas has developed an AI detector capable of identifying AI-generated content in academic papers with a remarkable accuracy rate of 99%.
This detector is specifically designed for academic writing and is one of the few available tools in the market for this purpose. The research team used 64 perspectives to create 128 articles using ChatGPT, which served as training data for the AI detector. The model successfully distinguished between human-created articles and AI-generated ones with 100% accuracy, and it identified specific paragraphs within the text with a 92% accuracy rate.
However, it should be noted that AI detectors have not proven to be entirely infallible, as there have been cases where false accusations of cheating with AI were made. Nonetheless, the development of accurate AI detectors for academic writing is a significant step toward addressing the challenges associated with AI-generated content.