I think we need to apply social pressure on two sides: against people who don’t assume interactions in good faith initially on the one hand, and against people who abuse the good-faith assumption on the other. Both need to change or be excluded from polite society (i.e., we block them & stop interacting with them entirely), because they enable each other.
When people don’t interact in good faith, they justify suspicion. When people act out of suspicion by default, they justify antisocial tactics that prey on that suspicious mindset.
Unfortunately, this is hard to maintain in a ‘flat’ social ecosystem (like twitter) where friends & strangers are the only two categorizations and therefore almost all possible interactions are with strangers. Trust can’t go more than a single hop there. Compare to the fediverse, where (while everyone can interact with each other directly) people associate themselves with home instances of ideally only hundreds of users, with admins in charge of enforcing instance rules (and therefore, a user on an instance with a good reputation is acting as ambassador of that instance).
The other thing we can do is be disciplined in fact-checking ourselves. The only way to shift a culture from underneath is to act as though you are already living in the culture you would like to see while protecting yourself from the worst aspects of the culture you’re still in. In this case, this means being louder than the people doing misrepresentation *about* their misrepresentations without engaging in call-out-style tactics.