4 ms·

They're required to spell out their moderation policy and provide details on how it's enforced, but the law does not appear to put constraints on the policy. So

They're required to spell out their moderation policy and provide details on how it's enforced, but the law does not appear to put constraints on the policy. So it would seem to be consistent with the law to have a policy like We moderate depending on our values and feelings as we experience them at the time, and we do not strive for consistency in these judgments and they may change depending on how well our moderators have slept the night before or how the air pressure affects their sinusitus. That would seem to comply with the law while enabling the social network to maintain maximum flexibility.

orra 3y ago
That's clearly not a real policy. You can't trick a court with such obvious deception. In particular, moderators have bosses, and nobody can believe moderators can do what the hell they want without their manager having a say. > while enabling the social network to maintain maximum flexibility. In other words, giving the employees absolute discretion isn't normally what you would consider giving the company (owners/directors) flexibility.
- adrr 3y ago
  That’s why they are fighting it. They don’t want to put that Musk is the moderation policy.
  - bluescrn 3y ago
    'It's a private company, they can ban/block/censor whoever/whatever they want' Pre-Musk, wasn't that always the standard standard response to criticism of activism-driven social media censorship?
    - tzs 3y ago
      So? That's still the standard response. All this bill is doing is saying that the company has to state what their rules for this are.
      - sebzim4500 3y ago
        Sounds like the rules are "We do whatever Musk wants. He's an unusual guy".
    - adrr 3y ago
      California just requires transparency of the policies. They just need to put in writing that free speech absolutist Musk is responsible for all the moderation of policies like censuring the account that tracked his plane with public data.
- rgbrenner 3y ago
  There's definitely at least one moderator at Twitter that uses that policy.
- michaelt 3y ago
  > That's clearly not a real policy. Of course not - a HN posts are hardly long enough to contain the real policy which would undoubtedly be many pages long. hirundo is posting a hypothetical, exaggerated example. But Twitter could have a policy that gave them great leeway to make arbitrary decisions, by stating things like: * We use automated systems, which ensure a timely response and let us stay on top of many millions of tweets per day. However, these automated systems occasionally err in both taking down reasonable content and leaving up unreasonable content. We are constantly improving these systems, but with x00 million tweets per day some errors are inevitable. * Bright line rules are not always possible in moderation. For example, we would generally not censor images of Michelangelo's David, the Venus de Milo, or napalm girl, despite a general policy of not showing genitals or female-presenting nipples. Similarly, a post might be parody, sarcasm, exaggeration for comic effect, or otherwise acceptable. For example, we might allow NWA's 1988 hit protest song "Fuck tha Police", or a tweet by Donald Trump threatening nuclear war with North Korea, despite a general prohibition on calls to violence. Decisions are made by our moderation teams on a case-by-case basis, following the vague subjective guidelines found in appendix A * We hire our moderation workforce from around the globe valuing diversity and multiple perspectives and timezones blah blah therefore details that may be obvious to some audiences may sometimes be overlooked by our moderation teams * The fact a tweet has been reported by a large number of people does not necessarily mean it will be blocked, as some more visible accounts may attract more flags proportional to their larger audience. Twitter also suffers from organised 'downvote rings' which systematically flag posts by their political opponents, hoping to reduce the reach of those posts. We may ignore such reports, when we detect them. * Content moderation is interlinked with spam, astroturfing and bot accounts. For example we may block posts about buying cheap viagra, despite it not being extreme content or hate speech. * Our automated systems may identify accounts as suspected bot or spam accounts based on heuristics that may not be obviously related to our moderation policies. Even a newly opened account that has never sent a single tweet or followed a single person may be blocked. * The ever changing spam environment, the ever-changing language used by Twitter users, and the cat-and-mouse game with people who'd like to not be moderated mean the guidance given to our moderators changes on a daily basis. This document is a snapshot and our actions in the past and in the future may be inconsistent with it. * We are committed to helping users stay safe and control their twitter experience, through options like Unfollow, Filter Notifications, Show Less Often, Mute, Block and Report. Expand with a few more pages of waffle and voilà, you have a policy that doesn't actually pin you down to taking any particular action in any particular situation.
cornholio 3y ago
Since the moderation practices that must be disclosed also cover things like hate speech or harassment and other generally illegal speech, then the platform can't simply answer in that fashion, because the court will issue a default judgment against the "social media company that has not made a reasonable, good faith attempt to comply", as per AB 587. Also, the "feelings" of the moderators imply human moderators exist and are the norm. This might not be true and you would be submitting false information for an AI-heavy moderation pipeline.
- sneak 3y ago
  Hate speech and most speech that falls under the category presently described by social media users in 2023 as "harassment" is decidedly not illegal, and is in fact protected expression. Threats are not protected, but the vaguely defined concept of "hate speech" that is not already-illegal direct threats of violence isn't really a thing. This seems to be a common misconception. If it were violent threats, which are illegal, people would just call it that. When people say "hate speech" they are attempting to promote censorship of otherwise-legal expression that most of society nevertheless finds repugnant.
  - Guvante 3y ago
    You are mixing many different viewpoints here and it is diluting whatever point you were making. Harassment has a legal definition and so a law referring to it is not "as described by social media users"... You then decide by fiat that the entire thing is pointless because of how you feel which ignores the very real problems on social media of exactly the kinds of messaging you claim this isn't about.
- edgyquant 3y ago
  “Hate speech,” is not a real thing or illegal under US law
  - JoeAltmaier 3y ago
    Wrong, and wrong. Plenty of definitions that make it a 'real thing'. And plenty of ways it's being enforced across the US.
_ea1k 3y ago
I don't think that would comply with the reporting requirements. And it sounds like their real concern is that it'd provide a way to use those reporting requirements, and the discretion as to what consitutes good faith compliance, to pressure the company into taking actions. There also appears to be a real free speech issue here that is at least somewhat related to litigation that has occurred regarding editorial freedom of the press.

orra 3y ago

That's clearly not a real policy. You can't trick a court with such obvious deception. In particular, moderators have bosses, and nobody can believe moderators can do what the hell they want without their manager having a say. > while enabling the social network to maintain maximum flexibility. In other words, giving the employees absolute discretion isn't normally what you would consider giving the company (owners/directors) flexibility.

adrr 3y ago
That’s why they are fighting it. They don’t want to put that Musk is the moderation policy.
- bluescrn 3y ago
  'It's a private company, they can ban/block/censor whoever/whatever they want' Pre-Musk, wasn't that always the standard standard response to criticism of activism-driven social media censorship?
  - tzs 3y ago
    So? That's still the standard response. All this bill is doing is saying that the company has to state what their rules for this are.
    - sebzim4500 3y ago
      Sounds like the rules are "We do whatever Musk wants. He's an unusual guy".
  - adrr 3y ago
    California just requires transparency of the policies. They just need to put in writing that free speech absolutist Musk is responsible for all the moderation of policies like censuring the account that tracked his plane with public data.
rgbrenner 3y ago
There's definitely at least one moderator at Twitter that uses that policy.
michaelt 3y ago
> That's clearly not a real policy. Of course not - a HN posts are hardly long enough to contain the real policy which would undoubtedly be many pages long. hirundo is posting a hypothetical, exaggerated example. But Twitter could have a policy that gave them great leeway to make arbitrary decisions, by stating things like: * We use automated systems, which ensure a timely response and let us stay on top of many millions of tweets per day. However, these automated systems occasionally err in both taking down reasonable content and leaving up unreasonable content. We are constantly improving these systems, but with x00 million tweets per day some errors are inevitable. * Bright line rules are not always possible in moderation. For example, we would generally not censor images of Michelangelo's David, the Venus de Milo, or napalm girl, despite a general policy of not showing genitals or female-presenting nipples. Similarly, a post might be parody, sarcasm, exaggeration for comic effect, or otherwise acceptable. For example, we might allow NWA's 1988 hit protest song "Fuck tha Police", or a tweet by Donald Trump threatening nuclear war with North Korea, despite a general prohibition on calls to violence. Decisions are made by our moderation teams on a case-by-case basis, following the vague subjective guidelines found in appendix A * We hire our moderation workforce from around the globe valuing diversity and multiple perspectives and timezones blah blah therefore details that may be obvious to some audiences may sometimes be overlooked by our moderation teams * The fact a tweet has been reported by a large number of people does not necessarily mean it will be blocked, as some more visible accounts may attract more flags proportional to their larger audience. Twitter also suffers from organised 'downvote rings' which systematically flag posts by their political opponents, hoping to reduce the reach of those posts. We may ignore such reports, when we detect them. * Content moderation is interlinked with spam, astroturfing and bot accounts. For example we may block posts about buying cheap viagra, despite it not being extreme content or hate speech. * Our automated systems may identify accounts as suspected bot or spam accounts based on heuristics that may not be obviously related to our moderation policies. Even a newly opened account that has never sent a single tweet or followed a single person may be blocked. * The ever changing spam environment, the ever-changing language used by Twitter users, and the cat-and-mouse game with people who'd like to not be moderated mean the guidance given to our moderators changes on a daily basis. This document is a snapshot and our actions in the past and in the future may be inconsistent with it. * We are committed to helping users stay safe and control their twitter experience, through options like Unfollow, Filter Notifications, Show Less Often, Mute, Block and Report. Expand with a few more pages of waffle and voilà, you have a policy that doesn't actually pin you down to taking any particular action in any particular situation.

cornholio 3y ago

Since the moderation practices that must be disclosed also cover things like hate speech or harassment and other generally illegal speech, then the platform can't simply answer in that fashion, because the court will issue a default judgment against the "social media company that has not made a reasonable, good faith attempt to comply", as per AB 587. Also, the "feelings" of the moderators imply human moderators exist and are the norm. This might not be true and you would be submitting false information for an AI-heavy moderation pipeline.

sneak 3y ago
Hate speech and most speech that falls under the category presently described by social media users in 2023 as "harassment" is decidedly not illegal, and is in fact protected expression. Threats are not protected, but the vaguely defined concept of "hate speech" that is not already-illegal direct threats of violence isn't really a thing. This seems to be a common misconception. If it were violent threats, which are illegal, people would just call it that. When people say "hate speech" they are attempting to promote censorship of otherwise-legal expression that most of society nevertheless finds repugnant.
- Guvante 3y ago
  You are mixing many different viewpoints here and it is diluting whatever point you were making. Harassment has a legal definition and so a law referring to it is not "as described by social media users"... You then decide by fiat that the entire thing is pointless because of how you feel which ignores the very real problems on social media of exactly the kinds of messaging you claim this isn't about.
edgyquant 3y ago
“Hate speech,” is not a real thing or illegal under US law
- JoeAltmaier 3y ago
  Wrong, and wrong. Plenty of definitions that make it a 'real thing'. And plenty of ways it's being enforced across the US.

_ea1k 3y ago

I don't think that would comply with the reporting requirements. And it sounds like their real concern is that it'd provide a way to use those reporting requirements, and the discretion as to what consitutes good faith compliance, to pressure the company into taking actions. There also appears to be a real free speech issue here that is at least somewhat related to litigation that has occurred regarding editorial freedom of the press.