I remember hearing some concerns with the tool before it was acquired by Mozilla, did they end up resolving those?

I assume this involves sending all text to Fakespot / Mozilla for analysis, which is concerning from a privacy standpoint. Their privacy policy seems to snarf many different categories of data, and they say explicitly that they may link data to your identity and use it for marketing purposes: https://www.fakespot.com/privacy-notice

AI detection tools flag things like the US constitution and the Bible as AI generated.

Which either means we are living in a simulation and the people running this joint used AI to write those documents or it means they fuckin suck at determining actual AI written text. I am leaning towards the latter.