I don’t think OpenAI really cares about false positives with text watermarking lol they just don’t want to lose paying customers

99.9 percent accuracy sounds like a lot, but imagine that one among 1,000 college papers was falsely labeled as cheating. That could lead to some unfortunate consequences for innocent students.

A [WSJ] survey of ChatGPT users showed that as many as 30 percent said they would stop using ChatGPT if its output was watermarked.

arstechnica.com