Google and Section 230: US court ruling could “turn the internet upside down”
wsj.com
wsj.com
On the one hand there's moderation to censor, or remove, offensive, dangerous, or disruptive content. I feel that section 230 does a fair job addressing this.
On the other hand there's recommendation engines often promoting outrageous content, putting things in front of everyone, in order to optimize for engagement and profit. Isn't this acting as a publisher, choosing what to promote, instead of showing people chronological, friend, or user selected feeds? These companies are in control over much of what is being presented to people.
There's been massacres committed in various countries because big tech companies promoted certain messages, based on their secret algorithms. I've been thinking that these algorithms promoting bad content (for profit) are a bigger deal unpromoted (obscure) bad content.
https://www.wsj.com/articles/google-says-supreme-court-rulin...
The issue was that in 1996, it was not yet clear whether ISPs, hosting companies, and web services counted as "publishers" or "distributors". Compuserve vs. Cubby Inc found that Compuserve was not liable for users' content because it didn't moderate (i.e. was a distributor), while Stratton Oakmont Inc vs. Prodigy Services found that it was liable because it did (i.e. was a publisher). Legislators felt that this setup a perverse incentive and so enacted 230 to that companies would not be penalized for good-faith attempts at moderation.
[1] https://en.wikipedia.org/wiki/Section_230#Background_and_pas...
That is incorrect. It was not liable because it didn't moderate (i.e., was a dumb-pipe distributor) and because it had no reason to know of the defamatory content.
That latter piece is fact-intensive and case-specific. Which means expensive litigation in every case. Which is not remotely the same as a liability shield.
https://en.wikipedia.org/wiki/Wireless:_The_Essential_Charle...
Section 230 means you can have user generated content that isn't perfectly moderated. The result of section 230 being abolished isn't some panacea of decentralized blogs and websites, its usenet flamewars and 4chan.
As say a blogger, you would only promote 3rd party content you approved of.
Lawmakers don't care about the technical details.
Unmoderated forums work fine when they're a small group of people with a vested interest in preserving the norms of the community. They fail quickly if they can't establish barriers against the outside world that prevent people with no skin in the game from hijacking the community for their own agendas. The modern Internet is much more similar to the latter than the former.
And I think a good bit of the problems on the internet today are because moderation is inadequate on these huge platforms, but they are passed off as safe places for everyone. It has blurred the line between trustworthy and untrustworthy, everything is now just ambiguously okay-ish.
Plaintiffs are saying that even though Google is protected when someone uploads a terrorist video, Google should _not_ be protected when Google itself _recommends_ a terrorist video.
On the one hand, yes you should be responsible your own editorial choices in the content you publish. On the other, legislative power to impose that responsibility should be leashed to the existing narrow categories of unprotected speech, under strict scrutiny.
obscenity
incitement
defamation
fraud
fighting words
true threats
speech integral to criminal conduct
child pornography
Argument #1:
> The lower courts have mistakenly interpreted “publisher” to have its everyday meaning, referring to an entity or person in the business of publishing, and have at times compounded that error by insisting that section 230(c)(1) applies to virtually any activity in which such a publisher might engage, including making recommendations. But “publisher” in section 230(c)(1) is used in the narrow sense drawn from defamation law. If section 230(c)(1) is properly so understood, the imposition of liability based on a recommendation would not in every instance treat the defendant as a publisher within the meaning of that provision.
Argument #2
> the content at issue must have been provided by “another information content provider,” not by the defendant itself. Recommendations may contain information from the defendant, such as a hyperlink with the URL of material the defendant hopes the user will download, or notifications of new postings the defendant hopes the user will find interesting. The Ninth Circuit erred in holding that URLs and notifications are not information within the meaning of section 230(c)(1).
Argument #3
> the Ninth and Second Circuits erred in holding that section 230(c)(1) protects a defendant if it sends to a user content which the user did not actually request. A defendant is acting as the provider of an “interactive computer service,” and thus within the scope of section 230(c)(1), when it is providing “access...to a computer server.” A computer functions as a “server,” as that term is used in section 230, only when it is providing to a user a file (such as text, or a video), which the user has actually requested, or is performing other tasks (such as a search) at the request of the user.
Hey, Google may win this (I am pretty sure they will). But this claim that a loss would turn the internet upside down is histrionics. No, if Google loses, the internet will still be just fine:
> Thus, although some practices that might be characterized as recommendations could satisfy all three elements of the section 230(c)(1) defense, others would not.
> Search engines are in two important respects different from social media sites. First, search engines only provide users with materials in response to requests from the users themselves, and thus necessarily function as providers of interactive computer services. Second, although search engines provide users with hyperlinks embedded with URLs, those URLs are created by the website where the material at issue is located, not by the search engine itself.
[1] https://www.supremecourt.gov/DocketPDF/21/21-1333/247780/202...
Regardless of the pithiness of the aphorism, if the result of its application is the destruction of a site like HN then it doesn't matter what the aphorism says - it's functionally a HN-destroying bomb. It's a deontological reversal. When someone points out the consequences of the action, the intentions necessarily become the realization of those consequences.
So why shouldn't they be liable? Because that liability is yoked directly to its destruction. They are inseparable.
It allows a community or website the ability to enforce against the most obviously problematic content while not forcing it to over censor against what might be reasonable discourse or simply emotionally heated or slightly factually incorrect things. Nor do they need to E.G. investigate if something potentially libelous is true or not; that becomes something for the other parties involved to work out on their own via traditional channels.
Because the former is nearly impossible to avoid and the later is nearly impossible to accomplish.
The more dogmatically you apply a simple sound bite solution, the worse complex problems become.
Also [2] is a good one for those who think that current platforms "overmoderate" content.
[1] https://www.techdirt.com/2020/06/23/hello-youve-been-referre...
[2] https://www.techdirt.com/2022/11/02/hey-elon-let-me-help-you...
Regarding [2], it is wise (necessary?) for any online platform to obey laws in places where they operate. But when I read about censoring spam or hate speech for instance, I can’t help but think of the scene in Game of Thrones where (forgive my poor recollection) a wildling tells Jon snow how lacking in freedom he is despite thinking he is “free”. Anyone advocating for unabridged (legal) free speech ought to be aware of the fact that there will be speech they do not want to see on their platforms if they get it. That is the price of freedom.