Why detectors flag formal writing as AI (the perplexity problem)

This has been on my mind lately and I wanted to get the community’s thoughts on it.

Here is what prompted this post – something happened in my class this week that made me realize how big this issue has become. I am not going to share specifics to protect privacy but the general situation is probably familiar to a lot of you.

I think we need to have a more honest conversation about this because the current approach is not working for anyone.

Anyone have additional resources or experiences to share? The more information we pool together the better.