When Should the Law Regulate Content Moderation?

Only when necessary to protect five basic internet rights.

Thanks to Eugene for inviting me to guest-blog this week about my new article, The Five Internet Rights. The article endeavors to answer the (internet) age-old question: When, if ever, should the law intervene into how private entities moderate lawful online user content?

This question has taken center stage as debates rage over the proper role of social media companies in policing online speech. For example, Twitter's decision to suppress the Hunter Biden laptop story alarmed many people about the power large platforms exercise, or could exercise, over the dissemination of news and, thus, over electoral outcomes. It also spurred Florida to pass a law that attempts to prevent social media companies from discriminating against "journalistic enterprises."

Similarly, social media companies' decision to prevent users from discussing the COVID "lab leak" theory—a decision they later reversed after the theory achieved mainstream status—caused many to wonder if social media companies may be hamstringing the search for truth by arrogating to themselves the power to determine what constitutes valid scientific inquiry. It too generated a legislative response—this time, from Texas, which went even further by prohibiting social media companies from discriminating against any users based on their viewpoints.

Still others fear that social media companies are being too lax in moderating user content. They believe that disinformation and hate speech pose a much greater danger to society, and even to free speech, than does private "censorship." These concerns have prompted other states like New York to consider legislation that would force social media companies to take down problematic user content or else face legal consequences.

But even if these concerns are valid—even if the content moderation practices of large online platforms present serious risks to society—is regulation the answer? Or would government involvement only introduce larger problems? And even if state intervention into private content moderation could improve matters, would the Constitution permit it? That too remains an open question after the Eleventh Circuit enjoined Florida's forced carriage law while the Fifth Circuit upheld Texas's similar law.

In the face of these thorny issues, it might seem impossible to answer the foundational question of whether the law should ever intervene in private content moderation. Still, I believe an answer (or at least a partial answer) can be found by examining the topic from a different angle.

When considering whether the state should intervene in private content moderation, much, if not all, of the analysis to date has focused on the actions of social media companies. That's not surprising, since most online speech these days occurs on social media, and social media companies make most content moderation decisions. But social media companies aren't the only game in town when it comes to limiting user speech. And far more concerning, I would argue, than whether Facebook permits some news story or scientific theory to be shared is the fact that content moderation is now moving deeper down the internet stack.

For example, after it was discovered that some of the January 6 rioters used Parler to amplify Donald Trump's "Stop the Steal" rhetoric, Amazon Web Services, a cloud computing provider, famously booted Parler from its servers, causing the website to go down for weeks. Other infrastructural providers, such as Cloudflare and Joyent, have similarly revoked hosting services for websites because those websites have permitted offensive (albeit lawful) user speech. Registrars like GoDaddy and Google have taken to revoking domain names associated with lawful websites whose viewpoints they oppose. And even internet service providers may be getting into the content moderation game by blocking their subscribers' access to websites for ideological reasons.

Still, most concerning of all was a development that received scarcely any attention in the press. Shortly after Parler managed to migrate to an alternate host, it went dark again, but this time for a different reason. After complaints reached LACNIC, one of the five regional internet registries responsible for managing the world's network identifiers, LACNIC revoked more than eight thousand IP addresses used by Parler and its new hosting provider, taking Parler offline once more. A year later, Ukraine sent a similar request to Europe's regional internet registry to revoke Russian IP addresses.

Given how foundational these core resources are to the operation of the internet—and the fact that a deplatformed website or user cannot simply build his or her own alternative Internet Protocol or Domain Name System—these developments point to a world in which it may soon be possible for private gatekeepers to exclude unpopular users, groups, or viewpoints from the internet altogether. I call this phenomenon viewpoint foreclosure.

Understanding viewpoint foreclosure provides the key to determining when and how the law should regulate content moderation. It suggests that intervention should start with basic viewpoint access—the right of all users to self-publish their lawfully expressed viewpoints on the public internet. It also suggests that whether the state should intervene when private intermediaries deny internet resources to users for ideological reasons should ultimately turn on whether users can realistically create substitute resources to stay online. As I'll explain when I unpack internet architecture, the law can ensure viewpoint access by guaranteeing five basic internet rights: the rights of connectivity (connecting to the internet), addressability (maintaining a publicly reachable IP address), nameability (maintaining a stable domain name), routability (having one's packets faithfully routed through intervening networks), and accessibility (not having one's users blocked from accessing one's content).

Put differently, when it comes to content moderation, the only regulation we clearly need (although we may indeed need more) is this set of five basic and irreducible rights. In this series, I'll show why that's the case by answering the following four questions:

1. Can the state regulate content moderation?
2. Can a controversial user really be kicked off the internet?
3. Should a website have the right to exist?
4. How can the state prevent viewpoint foreclosure?

Tomorrow, I'll tackle the first question.

Comments

[Frank Drackman]

What is this "Hunter Biden laptop story"

[Longtobefree]

You will find it in the dictionary as an example of "disinformation".
It never existed, and it's a bad thing it was discovered.

[Abdul Abulbul Amir]

Content moderation should be subject to regulation when on platforms that are effective monopolies.

[dwb68]

^this.

[Krayt]

Except...the whole reason we are here is because Congress threatened section 230, unless the companies played ball censoring.

You seek federal regulation to solve problems of federal interference?

I recall Mithras, in the game Sacrifice, "the god of Death invited you into his homeland, after you killed the god of Fire? Incredible."

[dwb68]

No: The reason we are here is monopoly technology companies.

[David Nieporent]

There are no monopoly technology companies.

[dwb68]

gaslighting 101

[Nige]

There are monopolistic practices, though.

[Rossami]

You think the Parler example above counts as a success story?

[Krayt]

Who go along with the powerful because...?

Oh, right. Threats by government.

[jb]

Wait, which of the technology companies are monopolies? Seems weird to say "all of the social media platform companies are monopolies" when they are (with the exception of Facebook/Instagram) distinct companies trying to compete with each other.

[David Nieporent]

Except…the whole reason we are here is because Congress threatened section 230, unless the companies played ball censoring.

Congress did no such thing.

[Sarcastr0]

You continue to have this appeal to incredulity about twitter being threatened.

You're wrong; your proof is just that you're super sure.

It's pretty lame how you love to repeat this ipse dixit.

[Rossami]

Which, I think, is the entire premise of the series. If a user "can realistically create substitute resources to stay online", it's not a monopoly and do what you will. But if the user cannot "realistically create substitute resources", that's a monopoly so keep your hands off.

[Sarcastr0]

If there are monopolies we should break them up, not just open the door to bonus regulation. This is outcome-tailored.

I am also not sure what the difference is between a monopoly and an effective monopoly.

I thought twitter had a monopoly, but people on this thread explained to me that they do not; a monopoly requires some structural advantage not just network effects.

[Jonathan Affleck]

I have a related question.

RNC v Google: How Does 47 U.S. Code § 230 Override § 202?

I have been puzzled about this question for years. § 230 contains no specific override of § 202. Why doesn't the federal government enforce § 202 nowadays? It has enforced § 202 in the past for a voice common carrier, for a telegraph message common carrier, for a telex message common carrier, and for an email message common carrier. I have never found a definition of telegraph in the US Code. The usual state definition seems to be: a system that transmits a message electrically by wire or by wireless means. Under this definition Google is probably a telegraph message common carrier as well as an email message common carrier. Google without doubt charges for its email service.

47 U.S. Code § 153 - Definitions (11) Common carrier is distinguished from (51) Telecommunications carrier.

47 U.S. Code § 202 - Discriminations and preferences
(a) Charges, services, etc.
It shall be unlawful for any common carrier to make any unjust or unreasonable discrimination in charges, practices, classifications, regulations, facilities, or services for or in connection with like communication service, directly or indirectly, by any means or device, or to make or give any undue or unreasonable preference or advantage to any particular person, class of persons, or locality, or to subject any particular person, class of persons, or locality to any undue or unreasonable prejudice or disadvantage.

(b) Charges or services included
Charges or services, whenever referred to in this chapter, include charges for, or services in connection with, the use of common carrier lines of communication, whether derived from wire or radio facilities, in chain broadcasting or incidental to radio communication of any kind.

(c) Penalty
Any carrier who knowingly violates the provisions of this section shall forfeit to the United States the sum of $6,000 for each such offense and $300 for each and every day of the continuance of such offense.

(June 19, 1934, ch. 652, title II, § 202, 48 Stat. 1070; Pub. L. 86–751, Sept. 13, 1960, 74 Stat. 888; Pub. L. 101–239, title III, § 3002(a), Dec. 19, 1989, 103 Stat. 2131.)

Judge Newsom referred 47 U.S. Code § 223(e) Defenses to assert that an Interactive Computer Service (ICS) is not a common carrier even though the statute says nothing about an ICS outside the context of a defense. Judge Newsom may have been overreading the statute, which seems to pertain specifically to demand dial telephone service, in order to apply a restriction on defenses outside of the context of Defenses.

The common carriage issue needs to be sorted out by SCOTUS after a duel between attorneys.

I don't know whether it still has any relevance any more, but the FCC has ruled that email is a common carriage service.

The FCC ruled in 1979 that Email was a common carriage service in Federal Communications Commission [CC Docket No. 79-6; FCC 79-465] Electronic Computer Originated Mail (ECOM); Proceeding Terminated, Memorandum Opinion and Order, Federal Register, Vol. 44, No. 181, Monday, September 17, 1979, Notices, pp 53788-53800. The services of Gmail and a 2023 social medium platform hardly differ from USPS ECOM and would probably have been considered common carriage services back then.

I was an internal expert consultant for this case. It never occurred to any of the attorneys that email is not a common carriage service.

When I read the definition of ISP in § 230, I don't see how Google could possibly fit into the definition,. Of course, I read the statute from the standpoint of a patent agent and consider "including specifically a service or system that provides access to the Internet" to be a limitation that Google does not meet because Google provides a service within the Internet and does not provide access to the Internet. To me normal statutory construction is essentially equivalent to Phillips standard interpretation.

Google would only be an ISP by broadest unreasonable interpretation of the limitation within the statute.

PS. Here is the relevant paragraph of the FCC ECOM opinion and order.

17. Not only is the proposed service “communications by wire or radio;” It is also a common carrier activity. As has often been noted, the statutory definition of common carrier is not helpful: “‘common carrier’ or ‘carrier’ means any person engaged as a common carrier for hire ***” 47 U.S.C. 153(h). Our Rules shed little additional light on the issue: “*** any person engaged In rendering communication service for hire to the public.” 47 CFR 21.1 Like our Rules and the language of the Act, Legislative history is also less than illuminating- the term was not intended to include “*** any person not a common carrier in the ordinary sense of the term.” Thus, whatever guidance we are to receive on the meaning of communications common carriage must come from judicial interpretations and comparisons of ECOM with existing communications common carrier services already regulated under the Act.

[USPS ECOM provided fee-for-service common carriage. A modern email service charges a fee, which is a combination of work-for-carriage (“eyes-on-a page”) with barter (collecting user information for monetization), and with a fee for storage.]

[David Nieporent]

Oh, god. He's back. After the Supreme Court shockingly did not grant his frivolous cert. petition, and after (I guess) he figured out that his lawsuit actually had been dismissed with prejudice, he was laying low for a while. But now he's back with the same nonsense as before.

When I read the definition of ISP in § 230,

…you're hallucinating, because ISP is not defined (or even mentioned!) in § 230.

I read the statute from the standpoint of a patent agent and consider “including specifically a service or system that provides access to the Internet” to be a limitation that Google does not meet

Maybe you should read it from the standpoint of a native English speaker and consider that the word "including" indicates that the following language is not a "limitation."

Google provides a service within the Internet and does not provide access to the Internet.

That is correct; Google does not provide access to the Internet. Google does, however, provide access to a computer server, which is all that's required for an ICS.

[Jonathan Affleck]

Obviously, it was a typo, and I meant ICS.

David Nieporent seems never to have taken a course in patent law.

In a patent claim, including, comprising, and consisting of, introduce limitations on the claimed invention.

In this case, including is followed by specifically and says that an ICS must include specifically a service or system that provides access to the Internet. Google email service does not provide such a service or system. Google email service is not an ICS according to Section 230.

Is an ISP an ICS? An ISP certainly does include a service or system that provides access to the Internet. An ISP certainly seems to be an ICS. An ISP definitely provides common carriage, and an ISP has at times been defined to be a telecommunications service Section 223 notwithstanding.

In 2015, the Obama Administration formally adopted open Internet rules to “compel internet openness” among ISPs. To strengthen the authority of these rules, FCC then reclassified broadband access as a “telecommunications service.” This reclassification brought ISPs under the purview of Title II of the Communications Act. FCC believed that ISPs were more akin to traditional telephone transmission and warranted extensive “common carrier” regulation. Such classification gave FCC power to ensure that ISPs treat all internet traffic the same regardless of source. Industry groups challenged these rules in court arguing, inter alia, that FCC lacks the authority to reclassify. The D.C. Circuit in U.S. Telecom Ass’n v. FCC (2016) upheld both the rules and FCC’s decision to reclassify broadband access as a common carrier service.

In 2017, the Trump Administration repealed the Obama-era rules and reclassification with the 2017 Repeal Order. Specifically, the FCC reversed the 2015 Title II Order to “restore broadband Internet access service to its Title I information service classification.” In support of its light-touch regulation, FCC argued that broadband access is an information service or, in the alternative, “inextricably interlinked” with information services. In other words, since ISPs provide a single unified information service, they squarely fit the definition of “an information service” of the 1996 Act.

An information service can be or can provide a common carriage service. It may not be a telecommunications service from one president to the next, but despite Section 223 the FCC could certainly apply Section 202 to an ISP, which is an ICS, even if it does not consider an ISP, which is an ICS, to provide a telecommunications service.

[David Nieporent]

David Nieporent seems never to have taken a course in patent law.

In a patent claim, including, comprising, and consisting of, introduce limitations on the claimed invention.

I have no reason to believe you are correct, given your track record, but I have indeed not taken a course in patent law, and for the sake of argument I will assume that in a patent, "including" is a limiting word. In ordinary English, however, and in statutory interpretation, "including" is not a limiting word.

In this case, including is followed by specifically and says that an ICS must include specifically a service or system that provides access to the Internet. Google email service does not provide such a service or system. Google email service is not an ICS according to Section 230.

No, it most certainly does not say anything remotely like that. Stick to patents, if you're actually right about them.

[Jonathan Affleck]

In patent claim interpretation, whether BRI or Phillips interpretation, including introduces a list of limitations on the claimed invention. Including is synonymous with comprising.

Interactive computer service

The term “interactive computer service” means any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides access to the Internet and such systems operated or services offered by libraries or educational institutions.

The "including" clause lists two limitations after specifically. One need only determine what the participle including modifies to interpret the meaning of Interactive computing service. Without the preceding comma "including" would modify the immediately preceding noun clause. With the comma, it modifies each of the nouns in the noun phrase "any information service, system, or access software provider". SCOTUS tells us that grammar and punctuation are important matters of law.

To tell the truth, if I were a patent examiner and read a claim with similar grammar and punctuation, I would have made an informality objection and compelled the applicant fix the claim, but Section 230 is enacted law. A Court must deal with it. A litigant must propose his interpretation of the meaning of Section 230 (f)(2), and the Court will make a judgment on the meaning.

[MollyGodiva]

Websites are not common carriers.

[Jonathan Affleck]

Many people, who have nor participated in a common carriage litigation, make vacuous assertions about common carriers.

California has expansive caselaw of common carriage as the RNC v Google order for dismissal points out on p. 17.

The Complaint also sufficiently alleges that Google provides its service for a profit. (Compl. ¶ 15.) In its Motion to Dismiss, Google argues that it does not charge for its email service Mot. at 12), but that is not what the statute requires. As the court observed in Squaw Valley, stores that use elevators and escalators to move patrons within a store are considered common carriers, even though they do not charge for the specific use of those devices. 2 Cal. App. 4th at 1508. It is simply enough that the stores profit “from the utilization of these devices to assist customers in shopping at the store.” Id. So too here, it is sufficient that Google profits off the email services it provides to its users.

How Does Google Search Service Belong to the Class of Common Carriers of Messages?

A search page result is a document, i.e, digital merchandise or digital literary property, which the search engine creates to a user specification (or to a machine-learned user preference). Google search delivers this document to the requesting or targeted user by means of message common carriage of digital literary property. The document includes a URL along with text. The situation is analogous to an online purchase from Amazon Fresh. A sales clerk assembles an order, and Amazon delivery transports the order to a destination by vehicular common carriage. It is incorrect to focus on specific URLs. Either a full HTML document is delivered to a user's computing device, or a representation of an HTML document is created by Google software on the user's computing device from data (usually JSON or BSON objects) transported by message common carriage from the search engine's server. The representation (DOM – Document Object Model) is used to display the search results on the display of the user's computing device. The Google user works ("eyes-on-a-page") or barters for common carriage. (Google collects valuable information from the user's computing device.) One pays either an explicit fee or an implicit fee for Amazon Delivery common carriage service. Google search service makes the product and does not gather together products that originate with other sources.

[MatthewSlyfield]

The RNC v Google lawsuit was about Gmail, not Google's search page.

[Jonathan Affleck]

I have read the case documents. I understand what the issues in RNC v. Google were.

I was trying to illustrate how to think about common carriage of messages in the context of the Internet by addressing a service, which is common carriage although far too many people seem unable to understand how this service provides common carriage of messages.

Google search service is at issue in State of Ohio v. Google (21-CV-H-06-0274). This case goes to trial in 2024.

[David Nieporent]

and Amazon delivery transports the order to a destination by vehicular common carriage

Once again: this is literally the exact opposite of common carriage.

[Jonathan Affleck]

You may argue with Massachusetts, which regulates Amazon Delivery via Massachusetts common carriage law.

[Rev. Arthur L. Kirkland]

How many installments before the “it is imperative to create as many safe spaces and special privileges for liars, bigots, and religious claimants as we can as society increasingly rejects conservative preferences” conclusion is revealed?

[epsilon given]

We know that only certian "bigots" are kept off the internet, but not all of them, because you and Democrats in general are still given free reign on the internet.

As for "rejecting" "conservative" preferences -- it will be a good day in America when Democrat Government fascists and their Big Tech allies are finally forced to respect the free speech of Americans!

[Rev. Arthur L. Kirkland]

If by using "free reign" you acknowledge that the liberal-libertarian mainstream has won (and conservatives have lost) the modern American culture war and will continue to shape our national progress, I commend your familiarity with the reality-based world.

If you meant free rein . . . ah, never mind.

[epsilon given]

Your eagerness to not only prop up, but to outright celebrate the descending curtain of Fascism of the American Government -- and to call it "winning the culture war" -- really sickens me!

You undoubtedly would have been the type to be on the "winning side of history" when the Wiemar Republic fell to the Nazis, and when the Czars fell to the Communists -- or, heck, when the Jacobans proclaimed a permanent reign of terror. It's really odd how all these "right sides of history" are so destructive of human life ... and how they eventually collapse, because there's only so much you can do to control the natural human yearning to be free.

[Rev. Arthur L. Kirkland]

The winning side of the modern American culture war is the side that prefers reason, inclusiveness, modernity, education, progress, science, and the reality-based world to ignorance, superstition, bigotry, insularity, backwardness, dogma, and pining for "good old days" that never existed.

The modern American (liberal-libertarian) mainstream, which continues to shape our national progress against the wishes of our vestigial conservatives, Republicans, and faux libertarians.

[epsilon given]

If what you say is true, then the Democrats would have died out after they lost the Civil War. Instead, they doubled down, and what's worse, they looked over to Russia and discovered that it's possible to enslave everyone, not just blacks, and they've been working hard to do so ever since.

And the only reason why they've made any progress in their efforts, is because of their appeal to emotion, their destruction of the education system, and their denunciation of logic as "white supremacy" -- the latter being a particular sore point of mathematicians like me.

[Rev. Arthur L. Kirkland]

Another disaffected right-wing misfit who can't recognize the glorious trajectory of a half-century of American progress, with more decades destined to follow as our nation continues to improve (more diverse, less religious, less rural, less bigoted, less backward).

[Nige]

If there miraculously came into being a mechanism and agency to effectively police the internet it should most certainly stick to actual laws being broken, agreed on via some highly unlikely international convention, and operate within a clearly defined set of rules and net users would be automatically granted civil rights the net police could not abridge. 'Moderation' would not be one of their jursidictions, except for cases of threats, harassment, identity theft, fraud and, fuck it, spam, make spam illegal.

[Jay Ash]

Those five rights and more? That was net neutrality circa 2003? And the graybeard internet devs at the time all loved it, but libertarians et. al., squashed it?

Oopsies now libertarians are crying foul when net neutrality principles are squashed along with the people and their messages.

[epsilon given]

Perhaps, being in software development and IT in general, I've lived a sheltered life, but in all the discussions of "net neutrality" I've seen, I don't recall ever seeing anyone discuss "The Five Internet Rights".

Indeed, if anything, the "net neutrality" being pushed would have forced ISPs to carry Netflix streaming while being forced to foot the bill for the extra equipment needed to handle such streaming -- and that has nothing to do with "viewpoint foreclosure", the concern of this series of articles.

[mydisplayname]

Amen! Different concept entirely.

[Rossami]

re: "That was net neutrality circa 2003"

No, not even close.

[SimonP]

Readers might note that this baby professor is teaching at the University of Tennessee, not Texas. In case there was any misimpression that he had an impressive credential.

As one has learned to expect of the VC, this post is under-theorized, asserting five basic “internet rights” but failing to explain their foundation or how they might interact with the property or “internet rights” of that class of diverse and functionally distinct service providers lumped together as mere “intermediaries.”

It’s interesting to note how the conclusions are baked into the premises here, too, by the stipulation that “viewpoint foreclosure” is measured relative to one’s access to resources to stay online. In conventional free speech discourse, for instance, one’s access to a particular form of communication, or a particularly effective distribution channel, is not of note. The fact that I cannot, for instance, easily set up my own TV channel or publish a widely-circulated newspaper does not mean that I have any “right” to publish any op-ed or broadcast any television show, or for that matter purchase ad space or time, in a given medium. Nor do I have any right to a cable, phone or internet connection from a service provider or my choice (even though those providers might otherwise be considered “common carriers” while those connections are made and services provided).

Suffice it to say, I have little hope that this series of posts will do much to address the theoretical shortcomings or limited defensibility of Eugene’s own take on these issues, despite likely being in broad agreement with it.

[David Nieporent]

There's the gratuitous nastiness we've so often come to expect from SimonP, along with the idiocy of claiming that a multi-day series of blog posts does not provide complete detail about the blogger's thesis on the first day.

[Rev. Arthur L. Kirkland]

At a blog that regularly sprays racial slurs and cultivates an incessant stream of multifaceted bigotry, this comment is singled out as "gratuitous nastiness."

Carry on, clingers.

(I don't expect Prof. Nugent to use a vile racial slur -- unless he's auditioning for a regular gig at this blog -- but I expect him to conclude that the internet must be kept safe for various types of conservative bigots, science-disdaining liars, and all religious claimants. Will he depart to any degree from this blog's predictable, partisan path?)

[SimonP]

Nothing about the OP’s outline for future posts suggests that he has any intention of addressing the theoretical shortcomings of his whole framework.

He hasn’t provided a thesis or conclusion yet, and I am not critiquing him for not having one. I am just pointing out that the terms of the debate he’s introducing here are already loaded. And like I said, that doesn’t bode well for what he intends to do with them.

[DaveM]

In conventional free speech discourse, for instance, one’s access to a particular form of communication, or a particularly effective distribution channel, is not of note.

If you were physically forbidden to speak, though, that lack of access to any form would be “of note”, would it not? Same here. To forbid your ability to connect online is necessarily to forbid your ability to speak online.

Or, said differently, "online" is an entire category of communication, not just a channel.

[Stephen Lathrop]

Nobody is forbidden by law to connect online. That is not the question this OP raises. It does raise two questions.

Primarily it raises the question whether a would-be contributor to internet discourse can be entitled by law to the services of the publisher of his choice. Secondarily, it raises the question whether if that same would-be contributor wishes to practice publishing activities himself, and thereby attempts to become a publisher, can he be legally barred from doing so?

The answer to the first question is that there is no such entitlement, and there cannot be, while 1A press freedom still stands.

The answer to the second question is that there is no bar to becoming your own publisher, but to do it you must learn what publishing activities are, and learn to practice them successfully.

[David Nieporent]

Primarily it raises the question whether a would-be contributor to internet discourse can be entitled by law to the services of the publisher of his choice

No, it doesn't. It's not talking about publishers.

[SimonP]

The OP mentions “content moderation” effected by social media platforms, cloud computing service providers, IP registrars, and internet service providers. Many of these service providers are not “publishers” in any meaningful sense, but whether social media platforms are “publishers” is at the very heart of the question, and is under rather intense dispute.

[David Nieporent]

Many of these service providers are not “publishers” in any meaningful sense, but whether social media platforms are “publishers” is at the very heart of the question, and is under rather intense dispute.

It may be the heart of the public debate, but the guest blogger's whole point is to reject the focus on social media companies in the debate, and to focus on the "infrastructure" companies instead.

[epsilon given]

Readers might note that hailing from one University or another doesn't matter if their arguments are solid -- that credentials shouldn't carry weight or grant authority -- and that even the best Universities employ idiot PhDs, and even the worst Universities occasionally support a brilliant PhD.

Readers might also note that, for all this talk of "not having access to TVs", the "baby professor" is talking about technology levels similar to telephones: servers, IP addresses, and domain names, which are necessary to host even a personal server.

Finally, readers who actually care about credentials might also note that even a "baby professor" has more credentials than an pseudononymous commentor with a funny name.

[Rev. Arthur L. Kirkland]

Finally, readers who actually care about credentials might also note that even a “baby professor” has more credentials than an pseudononymous commentor with a funny name.

Perhaps, but a professor in the American mainstream of legal thought likely wouldn't be publishing anything here, for several reasons.

[epsilon given]

Heh, another pseudononymous commentator with a funny name -- only this one has the distinction of happily siding with modern-day Fascism. Heck, celebrating it!

It's no wonder that you don't want to treat this scholar seriously.

[Rev. Arthur L. Kirkland]

If you genuinely intend to try to persuade anyone that scholars are to be found at this blog, you obviously haven't read Today In Supreme Court History.

Welcome, newbie with much to learn!

(While you're learning to scoff at Today In Supreme Court History, don't miss this blog's regular publication of vile racial slurs! You might find that a special treat, not often found anywhere else within spitting distance of modern legal academia.)

[epsilon given]

If I wanted to have a regular publication of vile racial slurs, all I'd have to do is ask you what you think about Clarence Thomas and Thomas Sowell -- I don't consistently get them from this blog, and when they pop up, I strongly suspect they are sock pockets on your side -- because most of them are. (They are probably your various sock puppets.)

As for the members of the blog not being scholars, just because you don't understand them, doesn't mean that they are stupid. It just means you have a lot to learn.

[Rev. Arthur L. Kirkland]

You must not be familiar with the proprietor's record concerning vile racial slurs. A Google-compatible device could readily illuminate that point for you.

This blog launches vile racial slurs roughly weekly. The other bigoted slurs -- against women, gays, Muslims, Jews, immigrants, etc. -- are launched even more often.

Other than that, though . . . great comment, dumbass!

[SimonP]

I am perhaps focusing on Nicholas’ credentials because they are being deliberately inflated/obscured. He was a Karsh Fellow at the University of Virginia, for instance - that’s basically an apprenticeship for law professors, the sort of thing offered to promising candidates, but no one the institution itself is expecting to hire on a full-time basis.

But, you know, tu quoque. I took a cheap shot at Nicholas’ puffery, but I offered a substantive critique, as well. You are right that a good argument can come from anyone, including a pseudonymous internet commenter. You haven’t addressed it.

[epsilon given]

And I, for one, don't make it a habit of paying attention to credentials. Being highly credentialed myself, credentials don't mean all that much to me.

And you out-and-out lie about my comment. I specifically called out your notion that we're not necessarily talking about other people's platforms -- that we're also talking about fundamental, low-level access on the same technological plane as telephones and email. I specifically pointed out this:

Readers might also note that, for all this talk of “not having access to TVs”, the “baby professor” is talking about technology levels similar to telephones: servers, IP addresses, and domain names, which are necessary to host even a personal server.

And you haven't refuted that.

[SimonP]

I haven’t attempted to “refute” this statement because (1) it’s true and (2) it’s not a relevant response to my comment. Put plainly: so what?

Every one of the “rights” described in the OP entails a compulsion on another person, whether we’re talking about distributing another person’s speech, hosting it, providing access to it, maintaining an IP registration for it, etc. Analogizing that framework to telephones, and hence to other “common carriers,” is a conclusion, not a premise, that requires at the very least clear statement, to say nothing of actual justification.

[epsilon given]

It's relevant because you're ignoring a major reason why this so-called "baby professor's" arguments hold weight: he specifically mentions basic infrastructure that's easily analogous to telephone communication, which is particularly protected by law. This is something you completely ignore.

Here's a little thought exercise for you: SIP (Session Initiation Protocol) is a fundamental standard for connecting two telephones over IP. What legal theory is there that simultaneously requires those two phones get connected if the connection is made via telephone company (regardless of who's trying to make that connection), yet it's perfectly reasonable to discriminate against one or both telephones, because one or both of the owners are guilty of "wrongthink", because the connections are being made over the internet instead (eg, via SIP and VoIP)? And if those same connections are required for telephone connections, what legal theory is there that limits the communication formats to voice only?

[SimonP]

Is a legal theory required? All that your thought example really describes are two technically distinct ways to make a phone call. They each operate by clearly distinguishable technical means, and you presume some explanation is required not to regulate them so that the user experience is functionally identical.

I don’t see any reason for why that should be. Sure, we could decide to impose rules on internet service providers do that SIP and VoIP telephone services act like traditional, common-carrier telephone services, and we may well have already done so. But that just reflects a policy decision and weighting of interests, not a self-evident fact about what it means to have a telephone.

Moreover, the thought experiment has an interesting way of undermining precisely the analogy you’re trying to draw. Because one thing that is clear about it is that “traditional” and SIP/VoIP telephone services are technically distinct in ways that requires specific regulation to make them operate similarly. We have to make the latter “act like” the former. That being the case, it’s clear that a simplistic assertion of an analogy between telephone services and the infrastructure of the internet skips quite a few steps. Which is the point I am making here.

[epsilon given]

Ok, now that I re-read your comment, I see that you have problems with the law in question itself -- that already puts those kinds of restrictions on telephone companies -- but that's not the question the professor is attempting to address, is it? He's taking the law (and the moral philosophy behind it) as de-facto reasonable, and is asking why the same reasoning doesn't apply to the internet.

You are free to make the case that telephone companies should be able to discriminate against anyone for any reason, but if you do that, you'll be addressing a particular tenant that, for better or for worse, has already been accepted by most of us as something very reasonable.

[SimonP]

I don’t know where you’re getting that I have “problems” with anything. I am just highlighting that the “internet rights” that Nicholas outlines are abstracting out the basic mechanics of making them real.

That we have made certain kinds of choices with respect to other common carriers is beside the point, largely because what we require of them differs from what would be required of the various service providers Nicholas lists in the OP. What is the telephonic equivalent of an IP registrar, for instance? Of a social media platform provider? Of a cloud computing service provider?

[DaveM]

Still, most concerning of all was a development that … LACNIC, one of the five regional internet registries responsible for managing the world’s network identifiers revoked more than eight thousand IP addresses used by Parler and its new hosting provider, taking Parler offline once more.

I did not know about the IP revocations, and I’m in IT. That is a Code Red for the Internet. It goes completely against the very structure of the Internet, which uses a layered design to separate concerns about handling transmissions. Just like Post Office is supposed to look only at the envelope to decide where to deliver a letter, not its contents, IP admins are supposed to assign addresses to all without favor, as long as they meet the technical requirements.

I knew about AWS (and threatened to take my business elsewhere because of it), but I see the rot has spread right down into the bones of the Internet.

I am going to be following this series very closely indeed. Forget the Googles of the world, this is truly alarming.

[Stephen Lathrop]

That was the part that caught my eye as problematic too.

[Jonathan Affleck]

I have thank Nicholas Nugent for his thorough article.

[ReaderY]

The state can obviously regulate “content moderation.” It prevents phone and mail companies from regulating the content of the messages they carry. Nobody questions the legality of doing so.

Its form of “content moderation regulation” is effectively property allocation legislation, declaring that the content is the property is and remains the content of the users who post it, and prohibiting carriage contracts proporting to provide otherwise. By doing so, “this content moderation regulation” merely protects users of common carriers from having their property taken and waylaid by the carrier.

Government could most certainly do the same with social media platforms.

It’s been argued that when a social media platform company moderates content, it is no different from when a self-administering user group dedicated to, say, cribbage does. But cribbage groups existed long before social media, and when they communicated by phone or mail, it was easy to distinguish their moderation from the carrier’s.

Common-carrier type regulations can easily be limited to large social media technology platforms, and independent user group administrators can easily be distinguished from agents of the platform. Claims they can’t be are nonsense.

Government can simply outlaw business models in which provision of the underlying technology platform services is dependent on agreeing to receive content moderation, advertising, etc., just as government currently prohibits letter carriers from opening users mail and inserting advertising or simply seizing mail it doesnmt like, and a company attempting to offer such a business model would be shut down as unlawful.

In other words, Government has every power and right to flat-out criminalize social media companies’ current operating model if it wants to

[Rossami]

The article doesn't ask "can", it asks "should".

[Stephen Lathrop]

Government can simply outlaw business models in which provision of the underlying technology platform services is dependent on agreeing to receive content moderation, advertising, etc., just as government currently prohibits letter carriers from opening users mail and inserting advertising or simply seizing mail it doesnmt like, and a company attempting to offer such a business model would be shut down as unlawful.

That is a terrible analogy. To take it at face value would authorize government to shut down book publishing—among many other types of publishing business. Some of those others practice activities less obviously related to your description, but all of them practice business models which are not in the least like those of common carriers. They also in fact rely on business models which 1A press freedom protects, so of course government cannot shut them down.

[David Nieporent]

Government can simply outlaw business models in which provision of the underlying technology platform services is dependent on agreeing to receive content moderation, advertising, etc., just as government currently prohibits letter carriers from opening users mail and inserting advertising or simply seizing mail it doesnmt like, and a company attempting to offer such a business model would be shut down as unlawful.

In other words, Government has every power and right to flat-out criminalize social media companies’ current operating model if it wants to

No, it doesn't. No more than it has the power to declare that newspapers must subsist off subscription revenue and cannot accept advertisements.

[MollyGodiva]

Private websites have a First Amendment right to decline to publish anything they dislike. It is a simple as that.

[Rev. Arthur L. Kirkland]

Not in Federalist Society America.

Which, like most of what conservatives (claim to) believe, is largely illusory.

[epsilon given]

Considering Federalist Society America is the America where the Bill of Rights in particular, and the Constitution in general, are recognized, I would have to agree that such things are illusory.

But that's not a particularly good thing.

[epsilon given]

It gets more complicated when you consider that,

(1) We aren't necessarily talking about websites -- Parler is demanding a right to IP addresses and domain names, fundamental tools needed to communicate on the Internet, and

(2) That the websites in question are acting in concert with government agents makes the removal a 1st Amendment violation. It's outright government censorship!

[David Nieporent]

(2) That the websites in question are acting in concert with government agents makes the removal a 1st Amendment violation. It’s outright government censorship!

Twitter (not saying that there's something unique about them; just picking one company as an exemplar) has the absolute right to moderate some particular content if it doesn't agree with that content. That not only does not violate the 1st amendment, but is in fact an exercise of Twitter's 1A rights.

Twitter does not lose that right because a government employee emails the company and says, "Hey, doesn't that content violate your rules? Shouldn't it be taken down?"

Twitter does not lose that right because a government employee emails the company and says, "This content bothers us. It's bad and we would like it removed."

The only way there is a "1A violation" is if a government employee emails Twitter and says, "We demand you take down this content or we will punish you," and the company removes the content when it otherwise wouldn't have. But in that case, the 1A violation is by the government, not by Twitter.

[Stephen Lathrop]

A fraught question implied by the OP is whether, "content moderation," differs in any significant way (legally or practically) from private editing prior to publication.

It has been problematic that Section 230 cleared the way for internet publishers to publish content without reading and editing it first. That enabled the business giantism which now vexes so many. But it also created an emergency need to implement editing after-the-fact-of-publishing—lest unsupervised publication of contributors' swill destroy the very business models which empowered all that publishing in the first place.

Of course, during the pre-230 publishing regime, some would-be contributors chaffed at editors' constraints on what they could say—and began to think of editors as censorious elitists. Naturally, when in the post-230 regime so-called, "content moderators," began taking down contributors' posts after they had actually been published—posts deemed libelous, for instance—that felt to the angry contributors even more like censorship than the prior editing.

However, post-publication moderation has proved disastrously less efficient—and far more permissive for the contributors than the prior editing. The disastrous part has been the damage inflicted not only on the targets of defamation, but also on the public life of the nation.

Which seems not to matter at all to the fans of a completely out-of-control internet. They want no editing whatever, pre-publication or post-publication. But the practice they attack is not really censorship, it is just editing practiced too late to do much good, or sometimes in an automated way prior to publication. And that remains a practice, however enfeebled, without which press freedom cannot continue as a viable way either to do business, or to support expensive freedom.

Repeal of Section 230, and a return to the former regime of near-universal editing prior to publication would better serve both the publishing industry and the public life of the nation. It would also give contributors convinced they had been censored less emotional reason to see it that way. They have only been edited by their publishers—a practice which the 1A has always protected, and still does.

[David Nieporent]

Blah blah blah I think newspaper publishers should be gatekeepers blah blah blah.

[ReaderY]

The delusion here is to believe that people send content to Facebook seeking its blessing, and for it to “punlish” it, in anything like the way people send letters to a newspaper or magazine editor. They do no such thing. There’s no analogy to writing a letter to the edditor to be published in an editorials section. People communicate with each other directly using Facebook as a simple communications technology platform. Anything else Facebook does merely gets in the way. The claim it is anything remotely like a newspaper is delusional.

[Stephen Lathrop]

Nieporent, yes, as rivals in competition with each other, each a gatekeeper of its own content. That is the private power the Press Freedom Clause was written to promote and to guard.

It is also the only public policy yet found with capacity to keep published expressive freedom safe from government meddling. It is capable to assure the maximum practical freedom for everyone interested in promoting expressive freedom through publishing.

To make that successful, public policy must be chosen to encourage profusion and diversity among a myriad of private publishers. That way, every strain of public opinion sufficiently popular to win an audience will likely also find a like-minded publisher to promote and publish it.

That ideal may not ever be entirely realized, of course. But experience has already proved it can be approached closely enough to excel every rival policy anyone has yet proposed. Nothing better has ever been seen, and what the nation has now under Section 230 is far worse. It seems almost everyone is unhappy with the status quo.

Promotion of diversity and profusion among private publishers is also by far the best policy to make entry into publishing activity a realistic choice for the largest number of people who might want to do that—so the notion that it promotes an elite kind of gatekeeping is off the mark.

Invention of the internet so radically lowered the cost of publishing activities that to do them is now within reach of even people of modest means. No more than a few thousand dollars is necessary to set up and begin business as a publisher.

Except for one thing. Section 230, which enabled unlimited publishing consolidation has put near-impassable practical barriers in the way of would-be private publishers hoping to get started. The advantages social media publishing giants enjoy in the marketplace for advertising sales all but preclude successful startups by would-be rivals.

That is one among a list of reasons why passage of Section 230 was a policy blunder which ought to be reversed. Other ill effects include:

• The baleful effect Section 230 has had on news gathering—at great cost to the nation's public life;

• Promotion of a politics of censorship, as political rivals jockey for government favor which they hope to wield to their own advantage through over-consolidated media giants;

• A practical license to commit defamation without much fear of consequence;

• And lately, the emergence of internet advocacy models seemingly based on capacity of those with the largest audiences to raise money fast enough to bid defiance to courts which attempt to intervene against frauds and other illegal abuses.

[ReaderY]

Nobody uses Facebook to “publish” their material. Facebook is simply a technology platform people use to communicate with each other. Facebook’s claims otherwise are delusional nonsense.

Nobody gives a shit about Facebook’s “curated content” either. The advertising and other tripe Facebook sends to people’s feeds as part of its revenue model simply gets in the way of use. People do their best to ignore it. They woild much rather it not be there.

Facebook is at bottom a common carrier. Its belief it is a publisher is nonsense.

[ReaderY]

Or as to paraphrase Lincoln, calling a communications carrier a newspaper or a publisher doesn’t make it so.

[Jonathan Affleck]

I used to have a mail subscription to Time magazine. The USPS was neither a publisher nor a distributor. The USPS is a common carrier.

[Stephen Lathrop]

ReaderY, I beat you to that famous Lincoln aphorism years ago, saying that calling a publisher a common carrier does not make it so. Unfortunately for both of us, citing Lincoln does nothing to define what is a publisher.

You must define publishing in the context of 1A press freedom. You cannot do it merely in the context of what would-be contributors want, or by the purely formal definitions lawyers resort to when legal cases which do not turn on questions of press freedom are being tried.

Your error is that you suppose the only activity relevant to a proper definition of publishing is to tailor activities to suit preferences of contributors—and you do not even seem to notice that what contributors chiefly want is to get their stuff published. That is something most contributors have for centuries been powerless to accomplish unassisted. Contributors have always been constrained by practicality to submit their stuff for the approval of businesses classed as publishers, precisely because those businesses operate and fund means to accomplish publication, and contributors typically do not.

Whatever the legal definition of publishing may be, its characteristics in the broader context of 1A press freedom have always been determined by a suite of activities most commonly used among publishers, and usually not used in other business models.

Those activities include (but are not limited to):

1. Assembling an audience for useful or expressive content.

2. Providing means to distribute useful or expressive content to members of that audience, who remain free to use or ignore the content at pleasure.

3. Curating the audience by selection of particular content the publisher chooses to put before the public—and thus to make the audience more or less inclusive according to the publisher's practical estimate of both its own expressive preferences, and its prospects to raise money to pay for publishing activities and make a profit. Note that ability to make private profit pay for publishing activities will always remain an indispensable bulwark of press freedom. Usually, no other source of resources to pay the costs of publishing activities can be found, except by too much burdening the publishers' liberty to choose content at pleasure—which is the very essence of press freedom.

4. In the course of curation, to vet the content to minimize the hazard of inflicting damage by defamations of third parties. And to further avoid myriad hazards of dissatisfying both members of the audience, and the public at large, by exposing them to non-defamatory falsehoods which are nevertheless disruptive of public life, such as election hoaxes, money raising scams in the names of spurious foreign princes, lies about public health emergencies, or racist attacks on disfavored minorities, to name only a few examples from an all-but-endless list. Note that freedom of any publisher to construct a different list also remains indispensable.

5. To provide advantaged outlet to the publisher's own expressive opinions, if that is what the publisher chooses to do. To promote that capability was among the chief goals sought to be accomplished by the press freedom clause of the 1A. It was by that means that the founders themselves had used publishing to make the American revolution a success, and to promote adoption of America's Constitution. Promotion of speech freedom opportunities for non-publisher contributors was another such goal prized by the founders, and likewise facilitated by press freedom.

6. To monetize the audience by making its concerted attention available to would-be advertisers, whose money pays the typical lion's share of most publishers' operating costs, and thus underwrites continued private practice of publishing activities.

That list defines key elements in a business model widely used to support publishing businesses, including newspapers, broadcasters, and social media platforms alike. You do not need to practice all the items on the list to be classed a publisher. But you can be assured that any business is a publisher if it does practice all of those activities as principal elements in its business model.

Which among the various social media giants does not practice that list of activities? Which common carriers do you know of which can say the same?

[David Nieporent]

Nobody gives a shit about Facebook’s “curated content” either. The advertising and other tripe Facebook sends to people’s feeds as part of its revenue model simply gets in the way of use. People do their best to ignore it. They woild much rather it not be there.

No, they wouldn't. If it were not there, they would have to pay for Facebook, and nobody wants to pay for Facebook.

Facebook is not a "technology platform." Well, obviously it is, but that's absurdly reductionist. If all people wanted was a way to communicate with each other, they'd still be using mailing lists. (I mean, electronic ones, not snail mail ones.) If you wanted to say that Facebook's _instant messenger_ service was a common carrier, that might have some merit. But that's not remotely what main Facebook is. Facebook is a content distributor, not a content deliverer.

[Jonathan Affleck]

Facebook is Email with a fancier user interface. There are a number of legal definitions of distributor, but Facebook does not pay for the content it transports as a bookseller would pay for the books it distributes/sells.

Facebook has bailment of user content, but a bookseller/book distributor does not have bailment of the books it sells.

[David Nieporent]

I mean, that last sentence is 180° backwards. And there is no such thing as “bailment” of information.

And the first sentence could be said only by someone who never used email. Or Facebook.

[Jonathan Affleck]

I have created email and social medium systems. I know what's on the inside.

A different UI does not change the service. Both Facebook and Gmail provide a service that carries a message electrically by wire or by wireless means.

If I have bailment of a physical book, I have bailment of information. I can have ownership of the physical book, or I can have a copyright to the information, which is contained in the book. The federal government long ago legislated that substrate type is irrelevant to an issue of bailment or of ownership (as long as the information is in some sense non-transitory).

The fact of bailment of digital information was long ago confirmed in the Dial-A-Porn cases.

[David Nieporent]

A different UI does not change the service.

It can. But in this case it's not a different UI; it's a different service. Nobody views "I'm going to post X on Facebook" and "I'm going to send someone an email" as interchangeable.

If I have bailment of a physical book, I have bailment of information.

No, you have a physical book. Merely possessing a physical book — or even owning it — gives you no rights to the information in the book. (That's not to say that one can't have rights to the information, but those rights do not arise because of the physical book.)

And again, a bailment of "information" makes no conceptual sense; you don't understand what a bailment is. The alleged patent law class you took is unhelpful to you here. Kind of like the way it was unhelpful in your frivolous lawsuits.

[Jonathan Affleck]

I worked with the AT&T legal department on the Dial-A-Porn cases. The question was simple: What relationship did AT&T or an RBOC have to a pornagraphic narrative that was hosted in the Bell network? AT&T or an RBOC had bailment and charged rent for warehousing. The analogy with a car in a parking garage is easy to understand. A parking garage may receive a limited right to drive the car within the parking garage.

Bailment is a legal relationship in common law, where the owner transfers physical possession of personal property for a time, but retains ownership. The owner who surrenders custody to a property is called the "bailor" and the individual who accepts the property is called a "bailee".

If I lend a book, which I own to another person, he has bailment of the loaned book. I may transfer my right to read this book to this other person.

[Sarcastr0]

It's gonna be a growth industry in rationalizing regulations to make people play with right-wing assholes.

Which, to be fair, only joins a bunch of other similar shops within academia on all sides of the political spectrum.

I'm also not one who thinks there is only one true constitutional proof, so lets see what kind of legal foundation the OP comes up with. Plenty of things I think are bad that the Constitution allows or even contemplates.