My off the cuff opinion, find a way to verify the people on the site are not bots and actually real. This is a hard problem with some expensive solutions, privacy implications, significant trade offs and real cost, however it may be worth trying to address. Imagine twitter (X) or facebook without the noise.

It's basically impossible. The sites that manage to be strict enough just end up with bots which are real phones in a building somewhere.

You just have to design so the bot's aren't relevant. The problem with Twitter, Facebook, and friends, is that they push the bot content at you, even if you don't follow them.

Yeah this would definitely be a unique feature.

I still don't think it would work now though. People don't trust social networks like they trusted Facebook in the olden days, and they never will again.

Require an email address, charge a dollar a year and have a reasonable report / takedown policy. That’d probably be 99% of the problem gone.

This is an impossible to solve problem, attackers have the advantage.

[dead]