I agree with this, but don't forget that interpretation biases will still come into play as the law is enforced — the biases of police, lawyers and judges. So it still makes sense to read others' interpretations of what this might mean in practice.
I've seen this regarding serious legislation like this or even something as mundane as Apple's app store "guidelines".
I may be misinterpreting your wording here (my apologies if I am), but I'd assume the opposite. I'd think people who operate in a realm where text becomes action executed by a machine designed wholly around faithful, reliable execution of text fed to it would come to learn the reality-defining power of rules.
In this sense, law is similar to code, but far easier to exploit.
I would disagree. I feel like programmers see logical contradictions or loop holes in laws and think that if they make the argument in court, the court will segfault and they will go free.
In reality, courts use more inductive reasoning than computers, and aren't as easily tricked.
Have you ever litigated?
Litigation isn’t about “hacking” the court. If that’s how a lawyer is selling themselves, you’re being taken for a ride.
Cases are about resolving novel ambiguities in the law. The vast majority of disputes never make it to court. The two sides lawyer up and one of them is advised that based on the facts the precedent is in the opponent’s favor. As such, settlement is advisable. In a minority of cases, precedent is mixed or not applicable—the facts and circumstances are truly novel with respect to the law. Given the law is finite and reality is infinite, this happens more often than you’d think.
Lawyers thus argue how the law should be extended. Remember, case law is law in common law countries. Judges opinions aren’t interpretations per se, but acts of rule making.
I see way too few computer engineer criminal masterminds to accept this hypothesis at face value. ;)
Applying to be a lawyer in and of itself can be a process of figuring out the correct paperwork to fill out and asking for a special exemption on a piece of missing information or a missed deadline.
Ah, but don't forget how often the code that gets written doesn't do exactly what the writer expected! Or is exploited by another party...
"For show," to me, implies you can ignore it and charge forward, bull-in-a-china-shop-style. That doesn't work in law or computers; naive invalid input gets rejected by the first-stage parser, and a court complaint completely ignorant of the law can get tossed by the clerk before it even sees a judge's desk. Rather, hacking is understanding and exploiting the consequences of, and nuances within, the rules.
So, quite ironically, you are doing precisely what you are talking about when you take the words "hacker news" too literally.
I don't think rulebook following coders are particularly interested in debating like this on HN.
All it requires is for other humans to accept your interpretation. That depends both on the text, your position in society, and that of the other person. The text will convince a few people when the difference in power is small. When the difference in position is large, no text will support the person in the subordinate position.
Lawyers pretend that this is not the case, and that they have mastered objective interpretation of the rules. It is unclear whether they say that because they are stupid, or because they think you are stupid. As programmers, who really are trained in objective interpretation of rules, it's laughable either way.
I try to avoid the error of confusing my lack of knowledge of the grammar of a system with the system itself being stupid. Law and its application have flaws, but it's not an "anything goes" system as you seem to be describing it here.
From my social experience in high school, the board game rules nerds were more likely to become programmers than lawyers. The worst of them went on to be a pro poker player.
Cone of ignorance expands further, faster than the cone of understanding.
Unlearning is even harder than learning, which contributes the durability of ignorance.
"You'd think /hackers/ would understand that rules are for show."
Ya. I'd like someone to explain cognitive certainty. The opposite of "strong opinions, loosely held". How is everyone else so sure of what they know? The only certainty I have is knowing that I'm probably wrong.
When man found Truth, a worried demon went to the Devil who nonchalantly said; "Eh, I'll get them to institutionalize it".
This disallows the common practice of open-ended moderation criteria such as "Be kind. Don't be snarky". Proposed section (c)(1)(b) removes the safe harbor except for moderation criteria on the list in proposed (c)(2)(a), which is: "obscene, lewd, lascivious, filthy, excessively violent, promoting terrorism or violent extremism, harassing, promoting self-harm, or unlawful"
> Provide Definition of Good Faith. Second, the Department proposes adding a statutory definition of "good faith" which would limit immunity for content moderation decisions to those done in accordance with plain and particular terms of service and accompanied by a reasonable explanation...
Open-ended moderation criteria such as "Be kind. Don't be snarky" are at risk of not being judged "particular" enough, which is a requirement of proposed section (g)(5)(a), meaning that sites with open-ended criteria could be judged as "not in good faith" according to this definition, and lose their safe harbor.
If you want to be deemed a public square type of space that is fine but you don't then get to impose arbitrary rules about what is said, that would mean it is not a public place and is in fact a private space being editorialized by your private rulings on what speech should be allowed.
This is a very good move and clarifies everything for everyone.
I don't see why this is a good move, other than opening up a bunch of sites to liability. And it's going to make the ToS even longer so that they can make it "particular". Who wins here exactly?
Section 5c is particularly problematic. It disallows shadow-bans if I'm reading correctly.
Section f9 is...suspicious.
Section d4 seems painful for small providers. You lose good samaritan status if there isn't a good way to contact you.
Section c2A is the important one. It basically means that a site cannot remove content unless it is objectively obscene or extreme. Section g5A suggests that you can have a ToS that explains your moderation decisions. But c2A seems to say that you aren't a good faith actor unless you use the objectively reasonable moderations standards defined by the law.
It's not clear that this would even have an effect since clearly you can moderate based on your ToS. A sports site can remove non-sports content even if it isn't obscene.
It seems like the goal would be to say that content moderators are biased in some way and drop section 230 protection based on that pretext, but that'd require a court to find that the moderation is being done inconsistently, and being done inconsistently intentionally. I have a feeling there'll be difficultly proving that.
I could be wrong.
Maybe it’s time for another step.
where so long as it's clearly defined in TOS and moderation policy you can remove it. E.g. fb has a more clear policy that's public that says no false facts that can cause harm.
I'm a 100% skeptical of this administration and the (run on rant ahead) insanely religious zealot by the name of Barr who is terrifying in trying to force his crazy beliefs on all of us whilst creating a monolithic fascist-esque unitary executive who's whim (or that who has his ear last..) controls virtually all aspects of our lives...
but this seams reasonable and seems like it would answer a lot of the problems we see complained about on HN all the time?
right now one of the top posts is about YouTube take down with no info/recourse.
The only thing I don’t like about Barr is it’s hard to see him curtailing the surveillance state, but frankly that applies too almost everyone in politics since those who speak out against the cia/nsa tend to get into unfortunate high-speed car accidents.
Barr himself is horrifically partisan, which as the AG of your entire country is scary as all hell. Just read his speech in November 2019 to The Federalist Society (which itself is horrifying, the societies goals are simply at odds with what the USA has claimed to be).
It's as partisan as it gets.
Or is that okay, because its in-line with your own opinions?
he thinks that religion is needed side-by-side with our individual liberty style government.
read this whole speech it's truly insane i can't believe someone so intelligent (and effective), who was given so much power by Trump, has these views it's truly scary
"But today – in the face of all the increasing pathologies – instead of addressing the underlying cause, we have the State in the role of alleviator of bad consequences. We call on the State to mitigate the social costs of personal misconduct and irresponsibility.
So the reaction to growing illegitimacy is not sexual responsibility, but abortion.
The reaction to drug addiction is safe injection sites.
The solution to the breakdown of the family is for the State to set itself up as the ersatz husband for single mothers and the ersatz father to their children.
The call comes for more and more social programs to deal with the wreckage. While we think we are solving problems, we are underwriting them.
We start with an untrammeled freedom and we end up as dependents of a coercive state on which we depend."
https://www.justice.gov/opa/speech/attorney-general-william-...
f9 is suspicious how? This document is not intended in anyway to prevent anti-trust actions, so why is a specific exclusion suspicious? There's a lot of public push right now to act on anti-trust issues for large tech companies, so there's no secret there.
5c... is more complex, and I'm not sure it means what you think it does.
c2A does seem particularly problematic to me in the context of trying to combat disinformation which is neither illegal nor obscene, nor violent.
Can you explain more, I'm not really seeing it...
Also how is disallowing shadowbanning problematic? I always never liked it.
If instead you outright ban them, they adjust or learn about your anti-spam algorithms and just keep going.
I'd say this administration has resonated with its base by treating bans as politically-motivated.
Facebook also regularly censors viral pro-trump content and project veritas has exposed this.
I wish it weren’t so, but it is. I’ve experienced it firsthand.
It's tough to take these at face value because we trust and believe in the things that define our ideology and take any challenge to our presentation of that as an attack on the underlying ideology.
Also this sort of stuff can lead to real world violence like when Facebook let violent vigilantes organize on their site it led to Kyle Rittenhouse killing two people.
I get to take down any flyers on my fence, even if I leave up the ones I like. Or I can vandalize the ones I dislike, maybe entirely changing their message. My fence.
No they're not; the government has already shown itself able and willing to do an end-run around the First Amendment by e.g. pressurising payment processors to refuse to do business with such websites.
> The owners of the shadow-banning websites should also have the right to decide what thoughts are exposed on their website.
> I get to take down any flyers on my fence, even if I leave up the ones I like. Or I can vandalize the ones I dislike, maybe entirely changing their message. My fence.
If you're hosting a private website you can privately decide what goes on it. If you're holding yourself out as a public communications provider and want the benefits of section 230, you're being granted special privileges by society and you need to hold up your end of the bargain by hosting the kind of uncomfortable discussion that society needs.
I cannot see how a change to a law that's proposed by the enforcement arm of government will actually protect people from abuses done by that same enforcement arm. If they act in bad faith now, why assume good faith will follow?
>you need to hold up your end of the bargain by hosting the kind of uncomfortable discussion that society needs.
I don't think Amazon's user reviews are where uncomfortable discussions need to happen. But I do think the government demanding what should be discussed on servers owned by private citizens is a clear violation of the First Amendment.
This seems like an argument that no government will ever reduce its own power, and so constitutional protections, due process etc. are all pointless. Even a single arm of the government is far from a monolith, and the boundaries of what government should and shouldn't do are always evolving. I'm sure this proposal isn't coming purely from the good of this administration's heart; part of it is public pressure, and part of it is the consideration that they may no longer be in power come November. But realpolitik is always a factor; good laws are still good laws.
> I don't think Amazon's user reviews are where uncomfortable discussions need to happen.
I'm sure a lot of vital IRL political discussion happens in the checkout queue at the supermarket (I'm sure that sounds like a joke, but I'm completely serious). As life moves online, we need a corresponding public sphere.
> But I do think the government demanding what should be discussed on servers owned by private citizens is a clear violation of the First Amendment.
Private citizens acting in their capacity as private citizens are still free to discuss whatever they like, or ban whatever discussion they like. If you want to have a capaciously moderated website that's fine, but such a website will not and cannot be a Section 230 public communications provider.
Type 1: Errors, ignorance, human foibles. Primarily annoying/asshole behavior.
Type 2: Malice and malformed content. Spam, propaganda, trolling.
I can argue against shadow bans for the first type. For the second type, any evidence of your operational method is a data point to break the moderation barrier and infect/manipulate users.
If a spammer knows its been banned, it switches over to another account. If a troll knows where your ban lines are, they come back and stay exactly on the edge to trigger someone or make them fall over the line.
Holy shit I did not expect this government to get something so nuanced and difficult so perfectly right.
Also clarification on unlawful and promotes terrorism language is a nice touch.
Kudos.
Thanks for the write up btw!
Can we please just do this one on its own either way? This has been a real problem online with companies like Cloudflare offering hosting to websites engaging in these areas. These three are explicitly illegal and yet sites that harbor this content, especially cyber-stalking sites like Kiwi Farms (47 U.S.C. § 223), are still somehow online.
There are two sides to this though. If you make a carve-out for ignorance you incentivize ignorance.
I think the argument is that if you’re not able to moderate your user-generated content at the most basic levels like running image hashes against the CP database then you shouldn’t be hosting it.
Then surely some minimum level of CP detection should be part of this section, right? If the requirements here are not defined well enough, then any company, from the smallest startup to a behemoth like FB, could be liable for some CP shared through the platform in a novel way that would have been impossible to detect.
Force providers to have a reporting system that feeds back a unique case code that can be quoted as evidence of knowledge. Then they have x days to investigate and respond.
I think there may be a light, or a deeper darkness, that comes out of this though. We may see a lot of investment in automation for catching this kind of content.
I'm not involved with HN, but it seems likely that many smaller venues esp ones that aren't big money makers, including most mailing lists and small sites like HN would be advised to discontinue operating if exposed to this kind of extraordinary liability over content which they had no knowledge of.
"Actual knowledge" should be your preferred approach to your concern... but nothing will probably solve your problem because extremely well funded platforms are substantially immune to the law in any case, just as they're immune to common decency.
No, they’re not social network type services like Facebook or Twitter, but... section 230 doesn’t discriminate between types of online services!
It's not all chatrooms and social media. Restricting the internet to be run by people able to manage their own websites would hurt. Ebay would have to manually review every account and listing. Good luck finding a user review website like Rotten Tomatoes. No more Straw Polls. No more GitHub.
It goes deeper than that, I think. Do VM hosting companies also rely on Section 230 immunity? Do ISPs? (They’re not Title II common carriers anymore!) Would providers of this nature be required to monitor what their users do as well?
I don’t believe either of these have been tested in court, but I think there’s at least the potential here to make it much, much harder to manage a website.
And automated systems have their own huge problems.
Fixed that for you.
You mean...Samaritan? The Samaritans were the Nazis of their day and the legend is about a Samaritan that rose above his race and did what was good - thus, the one Good Samaritan.
You wouldn’t say Bad Nazi. You just say Nazi and Good Nazi.
Someone already corrected you on facts, but I just want to say that the premise of "rising above his race" is nonsense to me. Why would you think this way?
1. Incentivizing Online Platforms to Address Illicit Content The first category of potential reforms is aimed at incentivizing platforms to address the growing amount of illicit content online, while preserving the core of Section 230’s immunity for defamation.
a. Bad Samaritan Carve-Out. First, the Department proposes denying Section 230 immunity to truly bad actors. The title of Section 230’s immunity provision—“Protection for ‘Good Samaritan’ Blocking and Screening of Offensive Material”—makes clear that Section 230 immunity is meant to incentivize and protect responsible online platforms. It therefore makes little sense to immunize from civil liability an online platform that purposefully facilitates or solicits third-party content or activity that would violate federal criminal law.
b. Carve-Outs for Child Abuse, Terrorism, and Cyber-Stalking. Second, the Department proposes exempting from immunity specific categories of claims that address particularly egregious content, including (1) child exploitation and sexual abuse, (2) terrorism, and (3) cyber-stalking. These targeted carve-outs would halt the over-expansion of Section 230 immunity and enable victims to seek civil redress in causes of action far afield from the original purpose of the statute.
c. Case-Specific Carve-outs for Actual Knowledge or Court Judgments. Third, the Department supports reforms to make clear that Section 230 immunity does not apply in a specific case where a platform had actual knowledge or notice that the third party content at issue violated federal criminal law or where the platform was provided with a court judgment that content is unlawful in any respect.
2. Clarifying Federal Government Enforcement Capabilities to Address Unlawful Content A second category reform would increase the ability of the government to protect citizens from harmful and illicit conduct. These reforms would make clear that the immunity provided by Section 230 does not apply to civil enforcement actions brought by the federal government. Civil enforcement by the federal government is an important complement to criminal prosecution.
3. Promoting Competition A third reform proposal is to clarify that federal antitrust claims are not covered by Section 230 immunity. Over time, the avenues for engaging in both online commerce and speech have concentrated in the hands of a few key players. It makes little sense to enable large online platforms (particularly dominant ones) to invoke Section 230 immunity in antitrust cases, where liability is based on harm to competition, not on third-party speech.
4. Promoting Open Discourse and Greater Transparency A fourth category of potential reforms is intended to clarify the text and original purpose of the statute in order to promote free and open discourse online and encourage greater transparency between platforms and users.
a. Replace Vague Terminology in (c)(2). First, the Department supports replacing the vague catch-all “otherwise objectionable” language in Section 230(c)(2) with “unlawful” and “promotes terrorism.” This reform would focus the broad blanket immunity for content moderation decisions on the core objective of Section 230—to reduce online content harmful to children—while limiting a platform's ability to remove content arbitrarily or in ways inconsistent with its terms or service simply by deeming it “objectionable.”
b. Provide Definition of Good Faith. Second, the Department proposes adding a statutory definition of “good faith,” which would limit immunity for content moderation decisions to those done in accordance with plain and particular terms of service and accompanied by a reasonable explanation, unless such notice would impede law enforcement or risk imminent harm to others. Clarifying the meaning of "good faith" should encourage platforms to be more transparent and accountable to their users, rather than hide behind blanket Section 230 protections.
c. Explicitly Overrule Stratton Oakmont to Avoid Moderator’s Dilemma. Third, the Department proposes clarifying that a platform’s removal of content pursuant to Section 230(c)(2) or consistent with its terms of service does not, on its own, render the platform a publisher or speaker for all other content on its service.
and
https://www.justice.gov/file/1319331/download
I highly recommend looking at the redline. It's approachable, and doesn't fall into the interpretation biases of the reporter.