Thread by @haroldfeld

Interestingly, the events of the last 24 hours highlight an issue I raised in Chapter 5 of “The Case For the Digital Platfirm Act.” Should laws or other legal obligations wrt content moderation apply only to dominant firms. /1

While we have much yet to investigate, it appears that most of the actual planning and organizing of the violence took place on smaller platforms such as Parler. Facebook and Twitter were used for organizing the general protests. /2

While one can argue that FB and Twitter allowed insurrectionists to gather the firewood, it was almost surely the smaller platforms where the real work of laying the fire and striking the match took place. /3

This highlights several factors. First, antitrust alone is not going to solve the problem of bad actors using social media platforms to organize violence. To the contrary, the proliferation of social media sites will increase the ability of bad actors to plan with each other. /4

This does not make antitrust useless, but we shouldn’t kid ourselves that if we just break up a few tech giants (or apply rules to a few dominant firms) we have done much to neutralize the threat of bad actors leveraging social media. /5

But we need a nuanced approach that reflects the different capabilities of larger v. Smaller platforms and how they are actually used by bad actors, as well as considering what harm we want to prevent. /6

This is why I have proposed a differentiation based on factors such as size and prospective harm, as well as respecting rights of free expression. In particular, I have urged: 1. The more the platfirm resembles a 1-to-1 telephone service, the more we treat it as such. /7

By contrast, the more it resembles a mass media broadcast, the greater the public interest obligation. /8

Second, certain obligations ought to be common to all platforms, particularly the obligation to report clearly defined “red flag” patterns to appropriate authorities, which can then seek a warrant. This is what we doin finance,for example. /9

Clearly it would not stop all bad acts, but such a system represents a reasonable and *feasible* balance of rights and legitimate safety concerns. /10