would you pay for a Chrome extension that highlights manipulation and propaganda tricks in texts on the Internet?

and who judges what is propaganda and what isn't?

A set of heuristics and a bit of machine learning. It won't judge, it will only highlight what it finds dangerous, letting you, the reader, to decide rest

it still doesn't answer the question.. who decides the criteria for highlighting certain parts instead of others?

Me, it's a set of heuristics that I am developing. There are simple like "emotional language" or "faulty generalisation", and more complex one, like flaws in logic

I'd be intrigued to try it, but I don't see myself paying for it. I'd be more inclined for an API that rates user-generated content for that kind of stuff, but even that is a stretch.