In web of trust, anyone could publicly certify who they know is a real person (i.e. validate a link from their id to another id). Then, if you received a message from someone, the system would find the path in the graph of real people you trust, to determine the trustworthiness of the source. So if the account is a bot, there would be no path from it to you in the trust graph.
The advantage is that everyone could supply their own subjective trustworthiness score, altering the graph. They could even publish it, so that other people could use trustworthiness assesment of accounts they personally trust.
The big issue with a system of web of trust is that it is too efficient, and just kills commercial advertising (and also propaganda). Because that is all about overcoming the natural web of trust that humans have.
This is propaganda, none of those supposed networks exists or were successful in anything and when the media do show some supposed accounts they don't have a lot of views. Please stop falling for this, your democracy sucks because the politicians suck and the people want change so they turn to extremist parties.