LLM-Assisted Proofreading

A study of 13 free AI detection tools found that a human-written text with LLM-assisted proofreading scored between 40% and 100% AI-generated, according to a paper published on SSRN by the author. The researcher tested seven human-written texts at four levels of LLM-assisted proofreading from three LLMs, confirming that free detection tools frequently misclassify lightly edited human writing as AI-generated.

A short one today. I applied to a fellowship last year, used an LLM to check my punctuation - wrote the disclaimer prominently as required - and later discovered the same text scored between 40% and 100% AI-generated on free detection tools. I didn’t let it go. When have I ever? So, I tested 13 free AI detection tools on seven human-written texts at four levels of LLM-assisted proofreading from three LLMs wiping sweat of forehead . The result: a published paper SSRN https://papers.ssrn.com/sol3/papers.cfm?abstract id=7073158 . It says what I suspected but needed to prove. Read it and let me know what you think.