That's the point, bad actors can circumvent the system if they feel threatened, but system can be exploited by gov/companies once rolled out globally to target any user, so we get something that may be not that effective against bad actors but poses great risk to be misused by gov/company in their own interests without users knowing. I've seen how an authoritarian gov in my country is targeting ppl bc they are uncomfortable for the system and this algorithm opens another potential vector of attack.