Comment by Icathian

> Private companies should not be the de facto moderators of free speech in our society. They are forced into that position by woefully inadequate governance by legal authorities operating multiple decades behind the current landscape. That's not what happened here. They made an appropriate decision.

It's not a difficult to say, "while we have no policies that restrict lawful content, we reserve the right to not service those who host and promulgate content that explicitly creates emergency threats to human life."

People and their companies aren't computers who have to allow everything to meet some absurd MVP product definition of false fairness.

> There were no good choices for Cloudflare here

100%. To me this isn't really about KF (which clearly sucks and should be offline, but through actual legal processes), this is a matter of, "When does internet infrastructure end and content moderation begin?" As I mentioned in a previous discussion[0], Cloudflare finds itself right at the blurred edge of this line, made more complicated by CF providing both hosting, which is generally seen as content, and DDOS mitigation, which is more ambiguous.

The same people who cheer this decision wouldn't be happy if, say, DNS servers refused to resolve mega.io because it hosts illegal pornography. Or if their ISP started blocking PTP or nyaa.si for copyright infringement. This is to say nothing, of course, of any suspect political interference in internet infrastructure, which we already see around the world[1].

[0] https://www.cnbc.com/2022/08/22/pakistans-former-pm-khan-say...

> I truly hope that those unsatisfied with this outcome (which I suspect will be literally everybody) can take this as an opportunity to go help pressure their respective governments to figure out what the hell should be done, systematically, about hate speech on the internet. It's only 25 years overdue at this point.

This is a very good takeaway, as it is a complex problem. But I think in the interim, it's perfectly fine for private companies with no legal obligation to keep sites like these operating to just choose not to do business with them.

  Private companies should not be 
  the de facto moderators 
  of free speech

Hate speech and organizing to harass and other IRL hate acts is not ‘free speech’. That was the major point.

I think Cloudlfare’s choice to block them is fine and CF was probably fine allowing their use of the service before, given the damage to their reputation they apparently considered acceptable.

Historically, you needed money or influence or both to make a “bad” (or in this case, actually bad) message widely available. What we’re seeing with Cloudflare and other companies choosing not to do business with some people is like a correction a bit back toward the past, after an hard swing toward unchecked, potentially widespread reach of speakers who wouldn’t have been heard much before.

>Cloudflare is choosing between "platforming the bad guys" and "censoring free speech".

Unless they provide hosting services, this seems a little distorted. Cloudflare is a DDoS protection service, not a platform. For nearly as long as there have been laws, there has been a general understanding that even the worst of us are entitled to the protection of the laws. Even Bill Cosby was entitled to his Fifth Amendment rights when he was given immunity for his infamous testimony. I don't see why Cloudflare's role should be seen differently; they have become the online anti-DDoS police, in the face of an Internet woefully under-equipped to manage such attacks.

Only in the case of the Daily Stormer, who deliberately turned Cloudflare's neutral role against them by saying "Cloudflare supports us", does there seem to be an exception, because they can't pretend to be truly bound by the law. But calling this "platforming" is basically playing into the hands of people running DDoS attacks.

Kiwifarms's hosts platform them. Cloudflare protected them. The difference is important. I don't know what happened exactly, so I can't comment on it, but I'm interested to find out what happened over the last two days.

On the one hand, I would have supported Cloudflare in continuing to provide service to Kiwifarms as someone not employed there if that was their conviction.

On the other hand, If I were the CEO, Owner, whatever of Cloudflare I would have cut ties with kiwifarms a long time ago on the grounds the site promotes truly immoral and reprehensible content and I wouldn't want any resources I control going toward helping them do so for my own conscience to be at ease.

Let's not obscure things by calling it an issue of "hate speech." That is an impermissible broadening. As they said, "hard cases make bad law." The only way to mitigate the badness is to make the decision as narrow as possible.

It's about illegal threats of violence. Those were against the law long before anyone ever used the term "hate speech."

I don’t think they’ve done well, and I think we can say that regardless of our opinion of this choice unless you think they should never deny service.

The mistake they’re making is this: they’re treating each event like a unicorn. They need to consider the overall decision making process. What are the inputs? What are the outputs? And they need to make these transparent.

The failure to do this results in the CEO publicly regretting previous ad hoc decisions. It’s also bad for the Internet. If you need to maintain the option to remove a customer — and you do — you need to be clear, consistent, and transparent.

It’s similar to ransomware decisions. You don’t want to make a decision about paying or not paying ransomware while you’re under pressure. Stress damages your ability to make rational decisions. Write a playbook and use it as your base for decision making.

This was not a free speech issue and I suspect that some of the attempts to reframe it that way are deliberately muddying the waters.

The issue at hand is that Cloudflare was providing material support to terrorists.

The site at the center of all this wasn't merely being critical of a group of people, it was being used to gather and disseminate personal information and coordinate acts of terrorism. Cloudflare meanwhile is not a public utility and had absolutely no obligation to provide services to terrorists; that was a smokescreen meant to deflect criticism of their decision to do so.

Free speech absolutists should really consider whether their argument is being strengthened or weakened by this specific case before hitching their ideological wagon to it.

Is it censoring free speech when the goal of the speech is to actively harm people? I’m not sure of any nation that has no caveats to their idea of free speech

Earlier they posted this image.

https://blog.cloudflare.com/content/images/2022/08/pasted-im...

That made sense to me. Basic services, transit, blocking ddos… little, if any moderation.

Hosting content, more moderation.

I might strongly disagree with someone and I sure as hell won’t host their BS, but I still think some basic level of rights/ services should be provided.

I think the internet is just irrecoverably broken in a way such that technical problems like DDoS or NN escalate to social problems. We should not even be having these discussions in the first place: It should be infeasible for attackers to conduct DDoS. It should be infeasible for ISPs to surveil their users. The internet as we know it was designed to facilitate communications between non-antagonistic peers, that design is no longer suitable for use by democratic society at large.

https://secushare.org/broken-internet

My main issue was this:

> Visitors to any of the Kiwifarms sites that use any of Cloudflare's services will see a Cloudflare block page and a link to this post.

Cloudflare was providing security from DDoS attacks. Then all the sudden they arbitrarily decided to hijack their domain. It would be one thing to stop providing protection. It’s another to say “no you see our content now”.

It would be like security at an event deciding to put in a band no one paid for. But still taking the money from the people hosting the event. The attendees are upset, the venue is upset, the original bands are upset.

Pretty sure that is a breach of contract. Feel free to drop them, but redirecting is wtf. Particularly, when they may be interfering with an investigation (as they said, cloudflare already took it upon themselves to involve law enforcement- who didn’t feel it necessary to shut it down).

>I truly hope that those unsatisfied with this outcome (which I suspect will be literally everybody) can take this as an opportunity to go help pressure their respective governments to figure out what the hell should be done, systematically, about hate speech on the internet. It's only 25 years overdue at this point.

I see what you mean and that sounds nice but how would that work? With the internet being international I can't imagine what could be done really. What KiwiFarm is hosting is already illegal in many jurisdictions I'm sure, but as long as the servers are hosted in some country with lax regulation (or a poorly implemented one) then what can be done at the state level?

Cloudflare has a really clear and seemingly mandatory option which is to just assert what we all know to be true: They can boot clients whenever they threaten their business. Everybody does this. There are near constant stories of users getting business-critical Google or Apple accounts suspended without warning or explanation. Those companies are ruthless about protecting themselves from liability and will err on the side of losing customers even when the violations aren't proven. Cloudflare wants to be seen as being above the fray and they just absolutely aren't. Anyone who thinks "it's just politics" is delusional. These are real people doing real things with real consequences.

It’s a fundamental issue though — there’s no “figuring it out” that a government can do that won’t either censor or facilitate. 25 years has been long enough to find tactical policy changes that make it easier, but there aren’t any, which is why nothing has happened. The choice we have to make is either de-shrine free speech above all else or entrench hatred, and it’s bogus that we haven’t picked the thing that doesn’t kill people yet.

> Given that they should never be in this position, Cloudflare is choosing between "platforming the bad guys" and "censoring free speech". They have navigated this imperfectly, but have done better than most would, I think.

They chose what's more convenient for them, as a private company, since the stock had a bad response after their previous statement.

They are, after all, a company that has to responds to their shareholders.

Why shouldn't we prefer private parties administer the free speech which is most fitting for their platform, as opposed to overly broad legislation at the national level by a government which doesn't appear like they'll catch up on tech within 5 years?

Pressuring government does not mean that the government will suddenly develop technical expertise. Even if the "right people" are voted into government at every level possible in a simultaneous magical moment, it would still take years for the government to develop its own internal consensus as to the state of problems and solutions & to slowly develop a workforce to administer technical policy. But one must question as to whether this is even in the cards for your respective nation.

In the meantime, if we prefer companies deal with the matter, people who don't like how things are done at least have the mere theoretical possibility of going it another way, assuming your market isn't so unhealthy as to only permit one entity (in which case you have a problem which a non-technical government may be able to deal with). The government issues monolithic force-backed policy, whereas the free market can create a diverse product ecosystem for different kinds of people.

And isn't the authority which is exercised by companies one which is ultimately quite fundamental — the freedom of association? The freedom to not have relations with those you don't want to talk to? Everyone should be free to yell their message on public property, but people should also be free to withdraw from each other if they no longer wish to be related. It is questionable to say that free speech must hinge on whether one party wishes to be related to another, especially when that other party has to maintain their services via expensive engineers.

> There were no good choices for Cloudflare here ...

I couldn't disagree more. Just cut off sites for organized harassment and nazis immediately. If you're able to, hand over any archives you have to law enforcement. Don't even talk about it. Just let such sites disappear one day. Don't give them any more attention.

Could you be breaching a contract? Maybe. But who cares? You think the KF owners are going to reveal themselves and sue? Let them.

This is what I find infuriating about the US media and many people in general: there's way too much effort spent on trying to appear neutral by bothsidesing every issue.

If CNN existed in 1938 Germany, after Kristallnacht [1], CNN talking heads would've gone on the air and said "sure a lot of Jews were killed, their homes and businesses ransacked and the authorities looked on without intervening but the perpetrators say they asked for it. Let's hear what this spokesman has to say."

[1]: https://en.wikipedia.org/wiki/Kristallnacht

>Private companies should not be the de facto moderators of free speech in our society. They are forced into that position by woefully inadequate governance by legal authorities operating multiple decades behind the current landscape.

When you have an algorithm that suggests things to people, you are a de facto publisher and whatever you do, you're choosing what to promote. In that case you have a responsibility to choose wisely, though it is a hard problem. Hard enough that in many cases algorithmic suggestions need to be avoided.

When you are a specialist with few customers there's no problem with picking who you work with.

When you provide infrastructure for large numbers of organizations though, you must be very hesitant to moderate who you serve, for many reasons. For the most part, if what you're serving doesn't break laws in jurisdictions you respect, they should be left to operate as they will. There is a narrow band around that of "maybe you should, maybe you shouldn't". There are real problems with expanding this to moderate the topic of shouting for the day.

I see a pattern that everything requiring international collaboration is very very slow to fix. From tax havens to climate change. Because you can hide out in another country or blame another country and so on.

The notion that government should be in charge of effectively eliminating speech we don't like so that private companies don't have to is _far worse_ than the current state of things.

The problem is who defines what is hate - don't trust a govt to make that - they are the last people I would trust.

We have no solution in this age - it was easy in the older days when consensus was reached within a village on what was bad for the community and you got either got tarred and feathered or thrown out.

I'm basically happy right now with private companies being the de facto moderator of speech because your alternative of government censorship is unacceptably risky to people's freedoms, and private companies are doing a decent enough job at it as is.

>what the hell should be done, systematically, about hate speech on the internet.

Nothing. It's words on the internet.

Clear and present danger is not protected under free speech, and that's exactly what the OP describes.

Hate speech is legal by design.

There is already illegal speech, not all speech is protected.

If they broke the law, charge them with crimes.

If they didn’t, ignore them.

"...pressure their respective governments to figure out what should be done, systematically, about hate speech on the internet."

Should be done by whom? If you mean "done by the government," then in the US at least the answer is clear.

My government should and cannot ever be trusted to moderate hate speech.

> Private companies should not be the de facto moderators of free speech in our society. They are forced into that position by woefully inadequate governance by legal authorities

I feel frustrated by this debate; do people really believe that having every speech question litigated in court is good for freedom of expression online?

My take has been for a while now that having multiple layers of enforcement for rules is a good thing because it allows for flexibility. It's bad for us to have only two categories for speech:

- morally obligated to host without question.

- will get you hauled in front of a judge.

The actual outcome of bigger companies like Cloudflare, Facebook, etc... pushing more of their moderation decisions onto the government is that the government will be doing a lot more moderating, and governments tend to be pretty clumsy about that, and court systems tend to be slow (for good reason, they have safety precautions because prosecuting someone is serious business), and laws tend to be very reactive and either overbroad or out-of-date, and they don't tend to take into account niche communities with special needs.

But beyond all of that, the law is also just a harsh thing to fall afoul of.

I just don't understand how someone can say, "make it easier to prosecute people for speech" and treat that like the pro-speech position. Isn't it better when communities and industries can have lower-stakes moderation decisions that aren't going to end up with someone being thrown in prison or fined? "The government should handle moderation" is exactly how we end up with bills like SESTA/FOSTA.

----

> Given that they should never be in this position

I also feel weird about this line. There are a lot of tech people who are fine with critical infrastructure being fully privatized, but draw a line at that infrastructure making its own decisions about moderation. If Cloudflare believes its services are de-facto public infrastructure, then why is Cloudflare a private company?

I feel like a lot of tech people want to have their cake and eat it too. They want to have a private company to be able to throw its weight behind technical decisions and infrastructure decisions and to shape the entire market, but they don't want any responsibility that might go along with that. My take is that if you don't feel responsible enough to be in the position that you're in, then get out of that position.

More and more, I realize that there is a difference between choosing not to abuse power and putting yourself in a position where you can't abuse power. Cloudflare makes a lot of excuses about how scared it is of abusing power, but what is it actually doing to reduce the amount of power it has over the Internet? If Cloudflare is saying that it shouldn't be making decisions about which services can get free CDNs, then that is tacitly saying that its specific CDNs and DDOS protection services are so powerful that they're essential to the modern web. If they're so powerful that Cloudflare wants to be completely hands-off about access to these services -- well, that prompts the question, "is it good for that kind of power regardless of the speech implications to be in the hands of private companies?"

Because Cloudflare has the ability to shape a lot more than just speech, it is in a privileged position to make decisions about core Internet infrastructure, and if its owners genuinely believe that they're not capable of making those decisions, then the irresponsible decision here is not in how they exercise that power, the irresponsible part is them holding onto that power and continuing to expand their marketshare and centralize that power even though they don't think they (or anyone else) is fit to wield it.

----

> Cloudflare is choosing between "platforming the bad guys" and "censoring free speech".

This point has gotten raised before by other people, but I just want to remind everyone that a nontrivial part of Cloudflare's business model is dedicated to censoring network requests. Cloudflare draws a line between "speech" and "abuse" every single time that it intercepts and blocks a DoS attack and every single time that it classifies IP ranges as dangerous or safe. It doesn't demand a court order in order to block clients from accessing a website if it thinks that those clients are contributing to a targeted attack. It doesn't rely on the government to tell it what is and isn't malicious web traffic.

A big part of the argument against Kiwi Farms was that the site wasn't just hateful or bigoted, it was actively abusing infrastructure and targeting individuals in the real world. I think compressing all of this down into a single "bad guys" category is oversimplifying the issue. I personally feel that Kiwi Farms has more in common with a malware or a DDOS-for-hire site than it does with a political site.

Of course, Cloudflare also provides services for multiple DoS-for-hire sites, and I have the same criticisms there. Is Cloudflare free-speech absolutist about this or not? Because if Cloudflare is arguing it has a moral duty to make sure that DoS sites stay online, then it's not immediately clear to me how it justifies its own business model. I honestly feel like there's a real lack of critical thought and real deliberation from Cloudflare recently about speech. I don't think the Kiwi Farms decision was a particularly complicated one, Cloudflare is a service that is in part dedicated to making it harder to knock people offline. That they don't seem to see any parallels between their services and their moderation decisions, and that they don't seem to realize that they are in fact in the business of censorship -- is concerning to me.

It feeds into the point above about why Cloudflare feels so comfortable being in charge of this much Internet traffic given its fears about censorship. Does Cloudflare not realize that it has a tremendous amount of power and is in fact directly shaping the kinds of expression and speech and services that people can build online regardless of what its moderation decisions are? Any company that's reaching the stage where they're scared about deplatforming a doxing site should be scared about a lot more of their powers than just moderation.

Let's say Cloudfare had not banned KF because they consider themselves a utility, what liability does Cloudfare really have?

> Cloudflare is choosing between "platforming the bad guys" and "censoring free speech"

Talking about free speech and censoring in this cases is what people that are ok with the harassment do. Don't. They should just have a normal TOS like everyone else and apply it. The freedom of speech should not be invoked in this case.

Yeah there was no winning move here. The pressure from both sides was too great

There was a better choice - do what they did, AND THEN JUSTIFY IT - show the court of public opinion redacted samples about why they took this action.

i think this is the sort of thing that exposes a general failure of capitalism.

the underlying assumption of capitalism is that competition works, and one company doesn't have the power to do something like censor a website, because competition will solve that. instances like this prove that to be wrong. cloudflare (and google, amazon, and most other big companies) get put in this position because regulators insist on pretending our economic system is pure capitalism. but in most cases, the big players are much more of a monopoly than anybody would like to think, and the forces of competition are a farce.

FWIW i think kiwifarms should be censored. but it sucks for cloudflare that they have to be the one to make that decision.

"hate speech" has no coherent definition, it's whatever the people with enough power to impose discourse limits don't like to be said.

[dead]

“Moderators of free speech” - an interesting idea.

>I truly hope that those unsatisfied with this outcome (which I suspect will be literally everybody) can take this as an opportunity to go help pressure their respective governments to figure out what the hell should be done, systematically, about hate speech on the internet. It's only 25 years overdue at this point.

I can unfortunatley see goverments doing a China style ID required for internet sites....

> figure out what the hell should be done, systematically, about hate speech on the internet.

I honestly wish there were some organization focused on the causes of hate speech rather than censoring hate speech.

What caused this? Why are Kiwifarm users so hateful? One does not just hate out of the blue, especially not to the degree of the actions they've taken (judging from their Wikipedia article).

Then there's the other end of this: To walk a thorny world, don't pave the world, wear sandals.

How can anyone be harassed online to end their life? Were there not enough settings, blocklists, and such to keep the harassment away? Were they unable to access the services that would have helped them better handle the harassment that did get through?