They aren't false positives though. They are legitimate positives.
Unfortunately most of the current reports are - the forum, signature managers needs to understand the difference between AI-generated posts and AI refined posts- the later is good use of AI.
This is what quillbot ( one of the advanced AI tools) has to say about the two and how false positive can be created.
Paraphrasing and grammar checking tools are ethical uses of AI writing technology. They help millions of non-native English speakers and writers refine their text and express their ideas fluently.
But most AI detectors can’t tell the difference between text that’s been created with generative AI tools—such as ChatGPT, Gemini, Claude, and others—and text that’s been refined using assistive tools. This leads to false positives and cultural bias.
https://quillbot.com/ai-content-detectorFurther furthermore, we should be encouraging people to learn English if they want to get paid for posting on an English language forum, instead of relying on tools that encourage laziness.
Using AI to refine their OWN opinion/posts to communicate with native English speakers fluently are also a way of learning. No non-native English posters deserves to be tagged AI spammer for using AI ethically. ( It's becoming a norms)