What your colleagues have to do is to pass their text through AI detectors, and if they get a result of 100% AI generated modify the result. The rules in the nutildah's thread are quite lenient on the subject tbh, they let you post texts with up to 70% AI generated.
The truth is that if you run an AI refined post on 5 different detector tools, you are certain to get different results; however, if you run an AI generated post on the same 5 detector tools, you will almost always get the same result--this is why you should use common sense and not base your decisions solely on what detector tools say.
This come to mind
Caution: Our AI Detector is advanced, but no detectors are 100% reliable, no matter what their accuracy scores claim. Never use AI detection alone to make decisions that could impact a person's career or academic standing.
Disclaimer on quillbot AI detector toolI admit that I used and explained there as well. Since it wasn't welcome, I leave using ChatGPT and right now just use QuillBot to fix grammar, not to rewrite the whole article or post. So anyone getting nailed should have a proper explanation. If someone writes the full reply through AI, then the post won't be compatible with the main post. However, I am against using a bot to write the whole post just for getting paid.
Fixing grammar and punctuation marks using AI refine tools is ethical use that's the whole point of my post.
I wrote somewhere on this thread or somewhere else that some of our phones keyboard are now AI powered and auto correct grammar mistakes before you can send out any post..