Concentrated Influence in the Internet’s Stack, Not Just Platforms, Creates Different Venues for Control
A couple of weeks ago, at the end of a piece about Substack’s alleged duty to moderate, I wondered how much of that demand came from its name-brand quality. Substack is more of a utility service than social media platforms, so if it were more subtle in its branding, I am not sure there would be so many people arguing that specific writers ought to be booted off. A user being removed from a platform is one thing, but having some crappy writer’s website disrupted at a DNS level would be more troubling.
So I was surprised to see a new reference from Geoffrey A. Fowler and Chris Alcantara at the Washington Post last week diving into some of those more infrastructural layers. It is an okay piece, but it is not without criticism. Let’s start with the premise in the headline:
Gatekeepers: These tech firms control what’s allowed online
That is only sort of true because there are laws that also restrict what is allowed online — and in other media — and laws that expressly permit what may be said. I bet you are thinking of one right now, so I’ll let Fowler and Alcantara take it from here:
A law known as Section 230 of the 1996 Communications Decency Act says “interactive computer services” — companies up and down the stack — cannot be held legally responsible for what others use their services to say. That provides them with a legal shield, with a few exceptions such as sex trafficking, but also gives companies the right to police content as they see fit.
Ah — so close. While Section 230 passes legal liability to individual users rather than platforms, it is the First Amendment that gives companies large and small the right to moderate as they would like.
So, we’re not off to a good start. But the questions raised by this article are worth considering:
Do companies have a responsibility to moderate content because they have the technical ability? Or does the fact that they could make the wrong calls mean they should hold back?
Fowler and Alcantara break the internet’s stack into five components, from those nearest the surface to those at the bottom:
Cloud computing and hosting providers
Content delivery networks, and security and payment services
Domain registrars (and DNS providers, but this article does not make the distinction clear)
Internet service providers
As you get further down in this stack, the power becomes more concentrated so the standard for moderation ought to be much greater. For instance, it does not seem to be controversial that platforms should make an effort to remove terrorists of any faction. But should ISPs make terrorist materials inaccessible? On its face, that is not a bad idea, but doing so could make it harder for researchers to do their jobs, and there is a whiff of the banned book emanating from this thought.
The fact is the entire internet stack is one private company on top of another. That means each gets to make its own decisions about how it wants to operate, and that will sometimes mean making hugely consequential decisions that reverberate upwards, often with little accountability aside from a brief public relations battle.
If you wanted to reduce the power of the companies at the bottom end of this chain, it would be a good idea to pass strong net neutrality regulations. A more radical idea would be to nationalize Tier 1 providers. I know that is a shocking suggestion for some, particularly in the U.S., but I would welcome it in Canada — though I imagine there are some unforeseen consequences and I will probably get emails calling me a communist.
Further up the stack, there is often more competition than the Post article indicates. For example, there are many more payment processors than are listed in the article, but they lack name recognition.1 One of the side effects of mainstream social media companies cracking down on disinformation campaigns is the increased popularity of other social networks. Unfortunately for people with humanity and a conscience, these are mostly platforms that have a permissive attitude towards white supremacists, which they laughably defend as a principled free speech stance. So, while I think the current use of Bitchute is abhorrent and I have no expectation that it will be serious competition for YouTube, it is at least an attempt at doing something different.