I won't worry about technicalities, because I have no idea what they are. But it is becoming abundantly clear that the social media companies have political power and are willing to wield it if they think that it is appropriate, similar to newspapers and television. It is reasonable that the regulations for social and non-social media converge.

This will leave an internet for those who can (i.e the big players).

Sorry, I think you actually do want a new law, not to modify this one. I have no idea how to define the websites that you want this to even apply to -- size? revenue? number of users? does twitter count? mastodon server instances? gab? reddit? texas instruments support forums? IRC servers?

This feels super arbitrary when almost every website to some extent republishes third party content.

(1) Someone has a long-established relationship with an internet service.

(2) That service makes an arbitrary and unexpected change to the nature of that relationship.

(3) That change causes material damage.

Then the site can be sued.

I don't think that would cover the current Trump business, they've been telegraphing for years that he would be banned eventually. But it might close the door to more damaging abuses of power where they strategic on timing and opens up some review of how clearly they are giving notice. Might help with the app store complaints.

That specific idea isn't going to be good - I've probably missed a bunch of relevant stuff - but maybe the idea can be salvaged and implemented with only mild damage to the internet ecosystem.

> Sorry, I think you actually do want a new law, not to modify this one.

There isn't really a difference, but sure. I don't care if changes are implemented as an amendment or a new bill.

That's not a good reason to DDoS them, though.

The DoJ proposal removes the catch-all and add "unlawful" and "promotes terrorism" to the list.

So if you moderate for a reason on that list, you keep the protection. If you moderate for other reasons, you lose it. And yes, I'm sure we'd see plenty of litigation on what constitutes harassment, terrorism, etc.

edit:

I personally think that the answer is to standardize terms-of-service documents and to make them into real legal statements that both the website visitor and owner can make claims regarding, and that those statements be registered with a federal agency. The standardization should be so tight that filing it should be a matter of picking from drop down menus on a form, and the implications of each choice should be common knowledge and taught in school. The form should be able to be filled out anonymously (as "owner of [X] site.")

If you run a site without having filed one, though, you would lose legal protections as a publisher based on modding behavior. If somebody filing suit can point out any decisions that have been defined by law as constituting a specific degree of editorial control (examples: spam-control, abuse-control, subject-control, position-on-a-subject-control, arbitrary-control, etc.) then they will be responsible for libel/tort claims as a site for individual posts. A site that has filed a terms-of-service will get the benefit of the doubt, and the ability to push liability onto completely onto posters or maybe even individual mods.

Those requirements would either be too onerous to run the walled gardens of Web 2.0 at scale and small players will have a profitable business opening again, or the big dogs would be forced to really bureaucratize or democratize, and small players will have a profitable business opening by providing them modding services at the forum level.

And sure enough, we only banned one person ever. We banned him because he made a series of scary threats against the release manager, including "I know where you live", but these threats were made off-list, it was a campaign of personal harassment because no one wanted this idiot's patches. Suppose we had forgotten to include threats of violence off-list in our expected set of rules? And remember, we were talking about our mailing lists, which are forums according to section 230. So, would we lose our section 230 protections under your proposal?

In other words, you want to create more regulations that websites have to abide by. All that does is favor the big players even more than they're favored now.

> Those requirements would either be too onerous to run the walled gardens of Web 2.0 at scale

No, they would be too onerous for any websites other than the huge walled gardens to meet. All you're doing is setting up another form of regulatory capture.

> or the big dogs would be forced to really bureaucratize or democratize

Or they'll just buy the regulations that favor them, the same way they do now.

Bluntly, "would need to be litigated to find out" isn't an acceptable answer given that the people giving this answer are the same people proposing the legislation.

Courts don't decide what the law is, they interpret it. You have to have an idea of what you want the law to be, you can't kick that off to the courts. They're trying to interpret your intention and the letter of the law you propose, not the other way around.

Let me predict how terror will end up being defined:

Promote storming the Capitol - No

Promote a free Palestine - Yes

Any fight against US oppression - No

Donald Trump's administration openly considered declaring BLM a terrorist group. I don't want someone like that deciding what does and doesn't count as neutral speech. That is a thousand times more dangerous to me than Facebook making the same decision.

It is startlingly naive for people not to understand that terrorism laws and 'neutral-speech' laws will be used against causes they agree with. Republicans in particular are kind of jumping the gun by encouraging the government to get involved with speech literally during a period where they no longer control the presidency, or the house, or the senate.

----

A) It's wildly abusable by corrupt governments, and both parties in the US believe that their opponents are corrupt. You don't even need to outright censor -- you can just classify one political group's speech as harassment and make your own speech un-moderatable.

And importantly, when you do that, it doesn't just apply to one platform like Facebook, it applies to everyone. It applies to independent forums that are owned by the other party. Do you really want to give either Republicans or Democrats the ability to basically at-will strip 230 protections from opposition-run forums whenever they control all three houses?

----

B) It places a burden on the courts that courts are not designed for. Courts interpret laws, they don't create them. The implication behind that is that the people making the laws have a general understanding of their own intent, and the Courts try to figure out what that intent is and what the implication of the letter of the law is. It is unreasonable for people to propose legislation that they themselves don't understand, and then ask the courts to retroactively come up with an interpretation and a set of guidelines that make the laws work. It's backwards design that will lead to a lot of unintended consequences and dumb, harmful rulings.

----

C) Many of these exceptions being proposed even in this comment section are themselves contentious. It is not clear to me, at all, why pornography should have fewer 1st Amendment protections than hate speech. Both are harmful to children, both can gross people out, both cause discussions to degrade, both are things that people don't want on their platforms. Yet, there is a general assumption that of course platforms will be able to filter lewd content, they just won't be able to filter low-key racists and nazis.

This is exactly the kind of content-favoritism that the 1st Amendment was designed to prevent. It is inherently problematic for the government to privilege certain categories of speech over others. The distinction between pornography and hate speech is one of the more blatant examples of hypocrisy, but the problem also comes up in more subtle ways when people talk about banning spam, as if comment spam, self-promotion, and off-topic discussion are some kind of clear categories that everyone agrees with -- they're not.

Reddit, Hackernews, and Facebook all have very different definitions of what self-promotion and spam are. Which definition is correct? Which one should be baked into law? And why? Why is it more reasonable for us to come up with an ad-hoc definition of spam than it is for us to come up with an ad-hoc definition of hate speech? Why should spam deserve less protection under the 1st Amendment?

----

D) Finally, baking these exceptions and moderating decisions into law creates an environment where we're constantly chipping away at the 1st Amendment. The way that the US works right now is that we have extremely broad legal protections, but that the systems built on top of those laws provide additional subjective moderation controlled by the free market, by individual choices, and by communities themselves. This is a good system because universal moderation decisions are practically impossible. In short, it gives us breathing room and allows us to acknowledge that moderation is both important and subjective.

When people tear away independent, private moderation decisions, the only thing that's left to protect people from harassment online is now the law. You lose that subjectivity and that acknowledgement that different forums need to be moderated differently.

People who advocate that private platforms shouldn't be able to make their own moderation decisions should not be surprised to see increased calls for the government to get involved in making certain speech illegal. Even if you could force Twitter or HN to turn into 4Chan, people aren't going to tolerate that. They are going to go to the government and ask the government and the courts to make more speech illegal -- because that's the only avenue they'll have left to protect themselves.

----

I have not seen any proposal, from anyone, not just in general in comments, but from political activists, from bloggers, from lawyers -- I have not seen one single proposal for Section 230 reform that addresses those problems. They all propose an incredibly ambiguous set of restrictions on moderation and then fall back on, "we'll work the details out later." But when you're talking about restricting someone's fundamental Right to Filter[0], the details really matter.

[0]: https://anewdigitalmanifesto.com/#right-to-filter

They're far from perfect, but is it better to give the power to control digital communications to a few corporations?

Newspaper publishing is opt in; that is, anything published they chose to publish.

Websites that allow third parties to post content on them is opt out; that is, anything published did so without initial moderation.

If a website operator posts their own statements, they can theoretically be found to be libelous. They can't be held accountable for posts by other people. Newspapers potentially can (though I've never seen a court case where a newspaper was sued for something in the Opinion section), but they -chose- to publish that item.

Realistically websites should be thought of as a public bulletin board. Should you be able to sue the person who put up the bulletin board, for content that was posted to the bulletin board by other people?

> When a US Army reservist found herself at the center of a conspiracy about the coronavirus earlier this year, her life was upended.

> Hoax peddlers on the internet falsely claimed that Maatje Benassi was somehow the world's COVID-19 patient zero. Over time, conspiracy theorists posted at least 70 videos across multiple YouTube channels claiming that Benassi had brought the virus into the world. Along with those videos came death threats, which Benassi and her husband, Matt, took seriously.

> But at first, the couple did not know how to respond. Trolls hiding behind aliases on the internet were almost impossible to find, and the Benassis could not sue YouTube for allowing the content to be posted because of a now-controversial law known as Section 230.

https://www.cbsnews.com/news/section-230-60-minutes-2021-01-...

https://revisionlegal.com/internet-law/defamation-attorney/i...

A) post video to YouTube for $0

b) contact lawyer to subpoena YT and then sue a jerk? $50,000

Worse, what if the defamer is able to hide their identity or from a jursdiction that doesn't care about an order from US courts? In that case, even paying for a lawyer won't help.

If you add black to a white background, you get black. If you remove white of a white background, you get black.

If the intention behind the action is the same, and the outcome is the same, should the legality of it hinge on the action?

Likewise, the New York Times is liable for the articles published by its own writers, but it bears no liability for the comments section.

The CDA draws a bright line between content "authored" by a firm and content "made available." In practice, that line is fuzzy.

As a hypothetical example, Twitter probably should face liability if it took a random tweet (say) accusing Bezos of pedophilia and made an editorial decision to promote that tweet to all its users, but it could still plausibly claim that it was just making the content available.

It's a complicated topic, and I don't know where the best balance lies.

That's what I meant, but you're right, I wasn't entirely clear. Thanks.

That's a protection that neither social media nor the NYT (for comments) would have without Section 230 if they do any moderation (at least according to Stratton Oakmont, Inc. v. Prodigy Services Co.)

Yes, the First Amendment does protect speech that gives false information. We had a recent HN thread on just this topic:

https://news.ycombinator.com/item?id=25695190

There already exists laws that act very similarly to what people are proposing.

Specifically these laws are called common carrier laws, as they apply to phone systems, for example.

Do you believe that the existing laws, that apply to phone networks, are tyrannical or worse or bad?

I don't think it is unreasonable to consider how our existing laws, that most people find reasonable, such as the ones that already apply to phone companies, could be updated to the modern era, so as to apply to newer communication networks.

I'm happy to let people I don't like have their own websites. I don't want to be forced to let them use mine to post about how I should die.

One company moderated their content, they were found liable for libel in user content in court. The other company did not moderate their content, they were not found liable for their user content. Anecdotally one of these services was seen as the good one and the other as the crappy one because of their moderation policies.

While social media didn't exist at the time in its current form, I'd argue this is still very relevant today. Without section 230 websites would only be able to moderate for illegal content, and there's a mountain of stuff you wouldn't want to see online that isn't technically illegal, and a lot of people who will try to put it in front of you if given the oppurtunity.

The concern is that absent Section 230, a provider needs to either be (almost) totally hands-off or they have to strictly moderate (which isn't really viable for the modern web and social media).