U.S. Senators propose limiting liability shield for social media platforms
reuters.com
reuters.com
It has been true for years now that the problems of the real world have spilled onto the internet. In its infancy, the internet was an "island of sanity," a (socially and intellectually) secure enclave from the real world.
In the 2010s, (and particularly the late 2010s) this problem accelerated, and now rather than merely bringing the problems of the real world to the net, the net is home to brand new problems which would have been impossible without the net.
It's no surprise that government wants to step in. It should be telling that both the left and right have plans for the government to regulate the internet. (and at least in my estimation, both of these plans are not well founded: either eliminate section 230, or enforce some kind of "fairness," or simply do more for censorship)
I honestly believe this is mostly a losing battle. Companies and the government, (and really people) will not have any reason to become less invested in the internet, and so the interests of individuals will be steamrolled. It's a totally unavoidable problem for most people, although a tech-savvy elite can avoid a lot of these problems.
What finally clicked for me is that power dynamics form a positive feedback loop that naturally compounds inequality. If you have a little more power, the natural thing do with that power is to coerce/force/encourage/persuade others to give you a little more. The longer you let that system iterate, the greater the power imbalance. This is why the majority of human history has had massive inequality.
Every now and then, an event happens that goes against that. For example, WWII destroyed massive amounts of physical infrastructure, so it was a force towards equality. Those with more to lose did lose more. (Heart-breakingly, the pandemic is the exact opposite, since COVID-19 doesn't harm material goods, just people.)
One kind of event is a rapid technological change. These work sort of like a rainstorm for power. All of a sudden, bits of empowerment from the new technology rain down semi-randomly onto humanity. Anyone with a cup to catch it gets some. The early Internet was like this.
But, eventually, the system iterates like it always does and eventually those with a bit more power use it to build the rain catchment systems and gutters to route that new power over to themselves. Sometimes, this gets routed to new powerful people who were the ones who caught a lot of the early rain. But the technology itself ceases to be a force for equality. This is where we are today with the Internet.
Just find another place on the internet with a barrier to entry that hasn’t been eliminated yet.
Back to Gopher and IRC I guess.
Yes, and happily so, for me at least. I use HN via RSS(except when making the odd comment), YouTube-DL via terminal, and the mainstream internet for ecommerce.
> American founders ... were not too keen on an unfettered discourse of the masses
I expect the founders' thinking about information sharing was largely based on their circumstances (printing presses, traditional mail service, town announcements, etc). Perhaps most founders considered them as immutable realities. Did any think of them as movable constraints?
Did any think about the downstream economic and political consequences of lowering the cost of communication? I don't know off the top of my head. Have you seen any evidence of this?
In all of these cases, Twitter and Facebook would argue that they don't have the manpower to quickly and thoroughly review every single interaction that occurs on their platform. And they don't have the luxury of being permissive in allowing free speech. And they would be right.
If you're afraid of online free speech being aggressively policed by tech companies, their customer-service reps, and their automated algorithms, then such proposals should absolutely scare you. You could very well see a 21st century where Mark Zuck becomes the next Rupert Murdoch.
I do not agree with your claim that "they would be both legally and morally justified in using draconian measures". Here's why:
1. Morality is different from legality, especially in this context
2. Draconian means "excessively harsh and severe". I don't think the bill if passed would provide protections for companies that are "excessively harsh or severe". These companies still operate in a broader context that includes market forces, public perception, and internal leadership in addition to regulation.
I sincerely hope that this is just political theater that never manifests into law. Section 230 is a blessing to the modern internet which goes wildly unappreciated, and is terribly misunderstood by the folks who benefit from it most.
In my opinion, a great lawyer will additionally discuss risks and the costs of mitigating them, particularly given a limited budget.
seems reasonable
>does not impair the enforcement of civil rights laws
what types of civil rights abuses were protected by section 230? I'm under the impression that saying/posting racial slurs isn't illegal.
> and does not bar wrongful-death actions.
what's the current bar for suing an establishment for wrongful death? if a bunch of radicals met at a bar and plotted to murder someone, would the bar be responsible if they didn't intervene?
Many people seem to believe that moderating platforms and banning accounts or otherwise enforcing terms of service for any content or behavior which isn't strictly illegal is a violation of their civil rights, and that this abuse is enabled by Section 230. They believe that any alteration of user submitted content makes the "platform" no longer a "platform" but a "publisher" and that they should be legally liable for that content.
The barkeeper might have a responsibility to listen in to the conversation and report if they are planning a crime, and the group is using the barkeeper's space to organize, but it's not like the barkeeper ever agreed for that specific group to use their space.
"[P]erfect knowledge"? They have knowledge of the text. You can even say they have perfectly accurate knowledge of the text. Do they have perfect comprehension and understanding of the text?
They have some comprehension and understanding in order to direct ads. But it is not perfect comprehension and understanding. And unless you're content with your apparently violent talk that's actually about a guild raid in WoW getting blasted by an AI, they need humans to ultimately decide what to do about content.
If the bar had a of method of grouping people together based on their interests, then did nothing about violent speech in the bar while promoting other violent interests and showing ads promoting body armour, then yes, the bar should be held responsible for acts of violence that come out of meetings in the bar.
By the way, Facebook is nothing like a bar.
The answer is yes. Yes, they would be liable.
Whether that’s a valid argument or not is up to you.
under what legal theory? A quick search says no. https://law.stackexchange.com/questions/3671/at-what-point-d...
What state do you live in?
It's even worse for tech companies, because of the way people use these services.
As a completely hypothetical example that I'm in no way condoning, let's say someone in a forum posted a comment like, "I hate n-words. We should off them all." The comments in the forum degenerate from there, and include posts of the form, "YY black church starts service at 10:30 on Sunday." and, " Oh yeah, you mean the N-word church at XXXX South Whatever Blvd in Charleston SC?"
I think you see where I'm going with this. Taken separately, each of those comments is either non-specific, or not threats at all. But taken together, it's a different story.
Now, imagine this hypothetical tech firm ran ads on their forums. It wouldn't even matter whether or not they ran ads on the forum in question, because the very fact that they decide where to run ads would make them a publisher and ineligible for section 230 protections.
Here is an idea for better oversight of big tech that does not involve more regulation: a deacentralized moderation system that replicates the us court system.
If something is flagged, then a jury made up of randomly selected users decide if the comment is inciting violence / etc.
If 1 user says it does not , then there is no ban/deletion etc. You can iterate many times off that into a system that is scalable, cost effective, descentralized, transparent, and just.
Note that it doesnt mean aws or gcp or cnn had to agree, it just means that it really resembled the process that we have for the us court system
In fact something well implemented would potentially threaten (and by definition, by attacked by) the twitters fb, aws , etc, because it would become a more reliable predictor of content moderation than what currently exists via ToS. It would effectively obsolte catchall ToS, and decidely be a vote against centralized control
We can likely agree that the words "well implemented" and "parler" do not belong in the same sentence
It depends on how the legislation is written, which will determine who makes various decisions and how the decisions are made.
Can you tell me about examples of legislation that you think have been done well? And what is your idea of 'well done'? Should I interpret your comment as meaning that you don't like arbitrary changes in political opinion -- that you think legislation should have deeper standards or principles? In the case of social media and speech, what do you think the standards should be?
Can you also give an example of legislation that has not been done well.
--- Given the regulation itself this is not a bipartisan effort (this is different from support bipartisan)
>>> Can you tell me about examples of legislation that you think have been done well
---I think the constitution is a sound example. Only the takings clause seems to have some pushback . Another hugely succesful act is U.S. Airline Deregulation Act. This is the only relatively recent act of congress that eliminated an entire US Agency (CAB), which itself makes it a rarity among the rare successes. Sherman act is one that everyone seems to come back to (and generally seems to be underpowered vs abused), the Voting Rights Act of 1965, Internet Tax Freedom Act, Personal Responsibility and Work Opportunity Reconciliation Act. Most acts restricting the power of the federal government tend to enjoy broad support over time.
>>>Can you also give an example of legislation that has not been done well.
---No child left behind ACT, Digital Millennium Copyright Act, Patriot Act, etc.
You can tie together almost all bad laws passed when they are (a) a reaction to specific events and circumstances (b) not bipartisan (c) usually sponsored by a very specific group. Such laws tend to not pass the test of time.
The reverse is also true, of course.
Would you like to finish the sentence? It is incomplete.
What is the intended purpose of such a definition? Is it meant to be explanatory? Predictive? Normative? Have you thought about the limitations of such a definition along these three categories?
Is this definition an attempt to find a neutral definition of legislative success? Or at least an attempt to find a definition that can be easily measured?
Can you provide detailed support for such a claim?
Next, would you be willing to talk about the complexities of gathering places, free speech, and who funds those gathering places? Would you be able to formulate one, two, or three basic theories (or models) of how these factors relate? Could you test them?
The reason I ask is this: I think there are some recurring patterns that connect these factors. I won't claim these patterns are universal, but I would expect to find useful patterns across many regions and many periods of time.
Having a tested, applicable model of how these are connected can help find the leverage points to target with regulation to achieve particular goals.
Caveat: what I've stated above is, more or less, a compact statement of how I think public policy should be made, at least for the most significant issues that we face.
And, if an issue is not deemed significant enough to warrant this kind of analysis, that's understandable. However, I don't think anyone should fool themselves into thinking that some 'obvious' action will also be verifiable *solution*. It might happen to seem effective to some people, but that doesn't make it an effective policy or decision -- very few of us are provably good at separating association from cause and effect.
This is a start, but the solution is to revoke 230 entirely.
Do you think that's an acceptable trade-off?
Just look at YouTube with kids. They’ve taken such a heavy handed approach to content moderation regarding kids content simply because of laws like COPPA. But then videos not involving kids get swept up into the mess because the algorithm decided their video is about kids. Even simple 10 second clips of Spongebob get flagged because of this.
Second, if websites shutting down is the tradeoff for harmful websites being able to be sued, yes, absolutely that's acceptable. Websites that propagate mass harm and disinformation need not exist, and their replacements will be better.
Also HN is high end marketing as an input for Y Combinator's founder program, if there was no business of Y Combinator funding it, it would not exist.
> all the new compliance cost
We don't need new compliance costs, we just need the DOJ and FTC to enforce our existing laws and break Google and Facebook up. They're already in violation of the law.
A big grey area seems like websites that give users tools to moderate content and let them handle it. Not sure what the courts would think of that.
I think it is also important to advocate for decentralized, federated communication channels. Increase the total number of possible points of failure, and they absorb the load in a more graceful manner.
Correction as of 12:57 pm EST: Based on a comment below, I see now that the link above is talking about a different bill from Senator Klobuchar with an anti-trust focus. She's been busy.
https://www.warner.senate.gov/public/_cache/files/4/f/4fa9c9...
And I do think platforms should be able to get sued and held liable for targeted violence against individuals that is incited on their platforms. The reuters article is light on information, but heres something that goes into a bit more detail.
https://royalexaminer.com/warner-announces-the-safe-tech-act...
Also, removing limiting liability doesn’t force companies to post things. That’s just 100% not what that is.
This part
>Under the SAFE TECH Act, the word "information" would be swapped out for the word "speech," narrowing the law and potentially erasing liability protections for a range of other illicit information-sharing that happens on online platforms.
I worry about. Who gets to decide what's classed as speech or not? What's the definition of 'speech'?
Will it end up being left to a judge to determine on a case by case basis?
In case you have not reviewed the 1A recently, please read this carefully: https://www.law.cornell.edu/wex/first_amendment
> The First Amendment of the United States Constitution protects the right to freedom of religion and freedom of expression from government interference. It prohibits any laws that establish a national religion, impede the free exercise of religion, abridge the freedom of speech, infringe upon the freedom of the press, interfere with the right to peaceably assemble, or prohibit citizens from petitioning for a governmental redress of grievances. It was adopted into the Bill of Rights in 1791. The Supreme Court interprets the extent of the protection afforded to these rights. The First Amendment has been interpreted by the Court as applying to the entire federal government even though it is only expressly applicable to Congress. Furthermore, the Court has interpreted the Due Process Clause of the Fourteenth Amendment as protecting the rights in the First Amendment from interference by state governments.
Please pay particular attention to whom and what 1A applies.
I’m not sure if you mean that the senators don’t understand. Or social media companies. Or us?
Government (in the US) regulates commercial speech all the time. Try broadcasting without an FCC license.
Government also regulated privately owned public places. Try banning people of a particular race from your shopping mall and see what happens. Shopping mall operators aren’t free to absolutely limit speech however they like, even though private.
I think it’s reasonable that government could treat social media companies as privately owned public spaces and add new regulations or requirements. I don’t think this would mean a constitutional amendment to revise 1A, but using the free speech principle for these regulations makes sense.
Of course, other countries have free speech laws and customs that have nothing to do with 1A.
To take just one example on this particular comment page:, I wonder if the person writing this comment understands 1A: "While this looks good on paper, it would severely hinder free speech on these platforms if not already."
While it is obvious that the comment doesn't say "First Amendment", that is not the end of the matter. When people talk, it is important to think about what they mean in a particular context. That comment very much relates to 1A, whether it uses a particular phrase or not, because it exists in this context of discussion about a bill talking about government regulation, which very much ties into 1A.
Let me circle back. My experience here on HN after trying to have substantive 1A discussions is, frankly, disappointing. Perhaps I thought that people smart enough to program computers would be somewhat good at unpacking logical arguments based on a complex legal history. I've found mixed evidence for this. Instead, I often see:
* Many people are simply unskilled communicators.
* Many people are unnecessarily pedantic -- meaning they make a tiny irrelevant point but miss the huge issues in play and don't connect with others as people
* Many people lack the tools or willingness to clarify, learn, and understand.
* In terms of subject matter, the basics of 1A are not well understood. Instead, dogmatic assertions tend to be used.
* FUD and slippery slope arguments are prevalent.
* Balanced discussions of pros and cons are few and far between.
* Even if the basics of 1A are understood, the authors often do not do a good job of formulating their questions in a way that signals their understanding and clarifies their question.
These are unfortunate patterns I've seen. This is why I shared some basic information to ground the conversation.
You may call me a critic, skeptic, or cynic. However, I think I'm being fairly accurate as to the quality of intellectual discussion around 1A on Hacker News.
So much of the 'legal' conversation I see here on HN would be laughed out of even an undergraduate law class. We can and should do better. It is not for lack of intellectual reasoning ability. Other things seem to be getting in the way: dogma, fixation on narrow things to the detriment of the broader context, ego, unexamined rigid beliefs, a lack of historical awareness, insular worldview, etc.
You are missing my point. If you re-read my comment carefully, you will see that I explained in some detail why I'm often disappointed in the quality of discussion here.
Only one part of my comments pertained to an undergraduate law class. I mentioned it not because I expect specialized knowledge of the law, but because I expect a certain amount of preparation before making strong claims. If you make a claim in a law class, you must be ready to support it. You are expected to know the history, arguments, and connection to the topic at hand.
I think many software developers can relate to the need to explain their claims in a work setting. For example, if a developer suggests that an organization switch from technology X to Y, they should expect to be asked why. So, it is not very different from the expectations of a undergraduate law school class.
I wouldn't say "absolutely" [1], but yes. At the risk of being too direct: I think we both know this is obvious. Do you have a finer point to make?
> There is absolutely an argument to be made that the scope of the 1st is too narrow to protect freedom of speech in the US.
I wouldn't say "absolutely" [1], but I tend to agree. This does raise a follow-up question: To what degree do you think free speech should be protected in the US? How do you balance it against other principles?
All of this said, I still stand by my commentary above, including (a) the unimpressive quality of argumentation and communication on HN; (b) a lack of understanding of how 1A and free speech interrelate; and (c) people here are not hopeless; they have the intellectual ability to do better -- if they only put a bit more effort into what they put out into the world.
[1] Are you aware of the pitfalls of using "absolutely"? I recommend reading https://www.dailywritingtips.com/absolutely/
I’ll frequently try to think about freedom of speech when trying to work out some community process or way to interact with people and if someone thought I was talking about 1A, I would feel bad for their confusion.
I’m not sure how to prevent this in my speech as if I said “free speech* *independent and not specifically 1A” every time it would probably be more confusing.
Yes, I understand this.
> Government also regulated privately owned public places. Try banning people of a particular race from your shopping mall and see what happens. ...
So far, this is not an issue of speech.
> ... Shopping mall operators aren’t free to absolutely limit speech however they like, even though private.
Ok, I see where you were going with it... I'm not a legal expert, much less in this area, but after skimming over [1] and [2] this appears to be a complex area.
[1]: https://en.wikipedia.org/wiki/Pruneyard_Shopping_Center_v._R...
[2]: https://www.ccim.com/cire-magazine/articles/states-speak-out...
Yes, this kind of argument seems to have a lot of traction right now, and I think there is precedent to build on, though -- again -- I'm far from an expert on it.
> I don’t think this would mean a constitutional amendment to revise 1A...
Agreed.
> ... but using the free speech principle for these regulations makes sense.
What do you mean by the "free speech principle"? I think 100 different people would probably have _at least_ 10 significantly different ideas of how free speech should work.
Good point, sorry, I said this like it was some simple concept that everyone understands in the same way.
I mean it in the terms of what the UN calls freedom of expression and I think is based on the Ancient Greek origin. [0]
That’s not absolute and there’s still debate, but, in my mind, it’s a philosophy that more communication on more topics is ultimately better for improving society and doing great things than less. An application that I think is relevant to this definition is Postel’s Law [1] to be liberal in what’s accepted and conservative in what’s said.
And I think part of this is to not try to have a rules based approach of specific banned topics because it ends up being circular, eternal, and nonproductive. So any efforts to get specific rules around this seem to harm groups.
[0] https://en.wikipedia.org/wiki/Freedom_of_speech [1] https://devopedia.org/postel-s-law
Seems pretty reasonable to me. It means if my system was exploited via an ad on Facebook[0], I can sue Facebook for showing me that ad.
[0] Situation is hypothetical, I don't actually use Facebook.
I hope you like spam.
Also, what you propose is equivalent to forcing you to pay thousands of dollars and wait weeks or months to take down an obnoxious sign that I put up in your yard. How workable does that sound to you?
Maybe you phrased it wrong, but the user cannot post death threats online without fear of consequences, it's just not the platform who the consequences fall upon.
A better analogy to a public location would be holding the city liable for not removing a death threat someone affixed into some city-owned monument.
That's only because the former is done anonymously and the latter can be easily be attributed to you, right? If a kid sent death threats to his classmates using his real-life facebook account, the police would be summoned too.
(If I were the communication tool, I'd add some term to my TOS that the user indemnifies me against losses due to speech the user made. So if I get sued because of something you said, I can come after you for the losses. Of course, this requires that you KYC . . . so anonymity might become harder in a post-S230 world. I sure wouldn't want to be someone running e.g. an anonymous far right or militia forum if I could be held liable for wrongful death or harassment claims.)
Just like in the real world, the venue is not responsible. Walmart is not responsible if someone shouts death threats inside. Walmart or other users/customers can call the police but it's not Walmart's job to tackle the guy. The really only difference is that local & state law enforcement is less inclined to enforce the law against online individuals given the time & energy required in addition the the threat being significantly less likely to be credible.
So I suppose it’s regulated in that it’s criminal and criminal actions are prohibited.
I’m not sure how you would fine Twitter because one of its users made a death threat, not how you would fine a McDonald’s because someone in their restaurant yelled a death threat. Or a megaphone manufacturer, etc etc.
[0] https://www.justice.gov/usao-wdnc/pr/federal-jury-convicts-b...
So it's technically correct to say that it's "just as illegal", but that's not the best kind of correct. While I'd be very interested in a directive that the FBI should take all violent threats online seriously, they don't have anywhere near that kind of resources. And I have a feeling that if they did start making numerous requests to unmask anonymous users, a lot of HN readers would oppose that as well.
If you report a death threat to the police and fbi it will be investigated. I think the problem is they have to be substantiated. If I just run around the town square saying “I’m going to kill you” that’s not actually a criminal death threat. So someone saying “#diaf” is not a death threat, it’s obnoxious but not a crime.
Fortunately murder rate is not correlated to internet death threats or we’d all be dead.
This doesn’t mean that platforms shouldn’t stop boorish behavior, but it’s not a crime and we probably don’t want the FBI investigating stuff that’s not a crime.
But when it is, when I think someone is going to kill me, I feel confident that I can call the law and they’ll do what they can.
[0] https://elkodaily.com/news/local/man-arrested-for-online-dea...
It has to be shown that there is both intent and the ability to carry it out before it starts entering the territory of illegal.
Which makes sense if you think about it. A 5 year old child threatening you with violence is not the same as a 25 year old man with a bat in their hand.
The reason threats of violence online are not illegal is because they have no ability to carry it out. This is actually a large reason why doxxing is such a big deal online. Suddenly, these people DO have the ability to carry out the threats.
If you REALLY want to go at it from that perspective, these companies leaking/selling PII should make them legally liable.
Ehh.
Slippery slope theory, if you don't want to be subject to these behaviors don't use social media. While a good bar tender will cut off patrons who are clearly drunk, if you then hop in your truck and run into a tree, you can't sue the bar.
Actually I don't think that's entirely true. I'm pretty sure there are laws in some places that if the bar knows a person is drunk and knows that they will drive, and does not prevent it, then the bar or the bartender (or both) can be held liable if the person then gets into an accident.
The Liquor Control Act prohibits alcoholic liquor permittees or their employees from selling or delivering alcohol to intoxicated persons (CGS § 30-86(b)(1)).
Although the statute does not define "intoxicated persons," the Connecticut Supreme Court held in 1937 that someone can conclude that a defendant is intoxicated if he or she is staggering and not able to run very well (State v. Katz, 122 Conn. 439).
Violations are punishable by up to a $1,000 fine, up to one year imprisonment, or both, for each offense
It is nearly impossible to engage in modern democracy without engaging in social media. I'm not saying this isn't a slippery slope, but when the only way to participate in our economy and government is online... well, there's a general public good element that needs to be considered by society. I'm not a big fan of government regulation, but the socials clearly aren't going to do it fairly.
I think you're using too broad a definition of "nearly."
Aside from HN, and the occasional write-only rant on Twitter, I do not use social media. And have never used it in any way to keep in touch with my government.
And I'm not alone. There are millions of people in the United States and hundreds of millions of people around the world who do not use social media.
https://www.google.com/search?q=bars+held+responsible+for+dr...
https://www.google.com/search?q=dram+shop+laws
"A commercial establishment may be held legally responsible for over-serving a visibly intoxicated person, or for serving alcohol to a minor, when that individual causes death or injury after leaving the establishment and causing: A motor vehicle accident, A pedestrian accident, Assault or other physical altercations, Other events leading to someone else's injury or death"
Regardless the impetus is on the individual to take an Uber to the bar if they plan on drinking.
This specific argument is weak, _and_ the form of the argument is weak. I suggest you look for better argumentation.
Your argument suggests that not 'using' social media frees a person from all of the effects of social media. That is oversimplified and untrue. It does not address the connectedness of people and culture, for one.
I'll use two examples to explain.
1. Your argument takes the same form as the following argument: "if you don't want to be run over by a drunk driver, don't drive". Tell that to a pedestrian who was killed by a drunk driver.
2. Social Media is not like Vegas. What you say on Social Media does not stay on Social Media. It has lots of spillover effects.