The correct way to deal with this is probably twofold: to have a variety of organizations which determine what is untrue, and to somehow figure how to make them seem credible.

Those organizations should probably operate independently of the companies actually implementing their blocking algorithms, and obviously of each other.

Then we need a diverse enough set of communications channels that each can make independent determinations how they filter communications on their platform, referencing the analysis of various organizations in the first paragraph.

This would maybe make for more controlled environments while not having universal censorship of controversial ideas.