One the one hand, there have been problems with Meta’s mostly automated moderation, and it is good if they are addressing that by hiring additional people for content review.
But it’s pretty clear that Meta’s updated Community Standards and moderation policies are designed to align with the incoming Trump administration. The updated content policies mostly focus on allowing content around gender, sexuality and immigration that was previously prohibited as hate speech.
Meta is also removing third party fact checking in the US in favor of Community Notes.
In a video statement, Zuckerberg calls out “governments and legacy media” that supposedly have been pushing censorship, and suggests that these changes are meant to push back against that and promote free expression. Right.
It’s not clear how bad this may be, and maybe it won’t be worse (or much worse), than the current state of moderation. But it has the potential to make Meta’s platforms much less pleasant for many, and even dangerous for some.
Today’s updates include:
Official Announcement: More Speech and Fewer Mistakes
The changes going into effect today can be found in Meta’s Community Standards Transparency Center.
Notably, the Hateful Conduct policy (formerly the Hate Speech policy) has been updated to allow more insulting language and content:
“People sometimes use sex- or gender-exclusive language when discussing access to spaces often limited by sex or gender, such as access to bathrooms, specific schools, specific military, law enforcement, or teaching roles, and health or support groups. Other times, they call for exclusion or use insulting language in the context of discussing political or religious topics, such as when discussing transgender rights, immigration, or homosexuality. Finally, sometimes people curse at a gender in the context of a romantic break-up. Our policies are designed to allow room for these types of speech.”
I suppose when members of Congress and popular X accounts are saying it, that means it's "mainstream".
Other changes:
They are now focusing the automated moderation on terrorism, child sexual exploitation, drugs, fraud and scams. Everything else will require user reports.
They are also adding more staff for reviewing content and requiring multiple reviews (including AI reviews) before content is taken down.
Will this be better? It depends on how the new systems work.
Instead they are launching X-style Community Notes in the US, which will require agreement between people with “a wide range of perspectives” to be posted.
If you are interested, you can sign up to contribute Community Notes on Facebook, Instagram, Threads.
It’s not clear that 3rd parties would ever be able to keep up with fact checking the flood of posts on Meta’s platforms, so maybe this was necessary. But a real drawback to Community Notes is that they are slow, and posts with misinformation can be seen by millions of people before a note appears.
This seems based on vibes, rather than any actual evidence that California teams “censor” more than Texas teams. And it’s insulting to both Californians and Texans.
The political content recommendations will be based on personalized signals from the posts you view and engage with. They will also expand the controls for how much political content you will see.
I think Meta’s definition of political content was always overly broad, not only including “government” and “elections”, but also “social topics that affect many people”. Social topics could include politicized issues like gender, immigration, climate change, and health policies.
But I’m not sure how you can have both friendly and positive communities, along with more recommendations of political social topics and less moderation of hateful content.
While he did not provide any specifics, this could include the EU’s Digital Services act that requires very large platforms to lessen the impact of misinformation.
- Updates to the Hateful Conduct (formerly Hate Speech) policy, that allows more insulting and hateful content, especially around homosexuality, transgender rights, and immigration.
- Focusing their automated policy filters on “illegal and high severity” violations.
- In the US, shifting from “biased” third party fact checking to X-like Community Notes.
- Moving the Trust & Safety enforcement team from California to Texas.
- Recommending political content from accounts you don’t follow.
- Working with the Trump administration to push back against foreign governments.
Official Announcement: More Speech and Fewer Mistakes
Policy changes allow more insulting and hateful content
Zuckerberg wrote that Meta will “simplify our content policies and remove restrictions on topics like immigration and gender that are out of touch with mainstream discourse.”The changes going into effect today can be found in Meta’s Community Standards Transparency Center.
Notably, the Hateful Conduct policy (formerly the Hate Speech policy) has been updated to allow more insulting language and content:
“People sometimes use sex- or gender-exclusive language when discussing access to spaces often limited by sex or gender, such as access to bathrooms, specific schools, specific military, law enforcement, or teaching roles, and health or support groups. Other times, they call for exclusion or use insulting language in the context of discussing political or religious topics, such as when discussing transgender rights, immigration, or homosexuality. Finally, sometimes people curse at a gender in the context of a romantic break-up. Our policies are designed to allow room for these types of speech.”
I suppose when members of Congress and popular X accounts are saying it, that means it's "mainstream".
Other changes:
- Removed the policy that prohibited comparing people to inanimate objects (with specific examples of women as household objects or property or objects in general; or referring transgender or non-binary people as “it”).
- Removed the prohibition on denying the existence of specific protected characteristics.
- They explicitly allow “allegations of mental illness or abnormality when based on gender or sexual orientation, given political and religious discourse about transgenderism and homosexuality and common non-serious usage of words like “weird.””
- They allow “content arguing for gender-based limitations of military, law enforcement, and teaching jobs. We also allow the same content based on sexual orientation, when the content is based on religious beliefs.”They allow expressions of hate (“I despise …”), expressions of dismissal (“I don’t respect…”), expressions of repulsion (vile, disgusting, yuck), and most profanity meant to insult.
- There also used to be additional restrictions for advertising and paid content, and those have been removed.
Focus filters on “illegal and high severity” violations
Meta has been criticized from both the left and the right for poor moderation by their automated systems. The problem was not so much the actual policies, but that the automoderation was not able to understand context.They are now focusing the automated moderation on terrorism, child sexual exploitation, drugs, fraud and scams. Everything else will require user reports.
They are also adding more staff for reviewing content and requiring multiple reviews (including AI reviews) before content is taken down.
Will this be better? It depends on how the new systems work.
Shift from fact checking to content notes
Meta’s started a 3rd party fact checking program in the US in 2016. Meta now says there has been “too much content being fact checked that people would understand to be legitimate political speech and debate.”Instead they are launching X-style Community Notes in the US, which will require agreement between people with “a wide range of perspectives” to be posted.
If you are interested, you can sign up to contribute Community Notes on Facebook, Instagram, Threads.
It’s not clear that 3rd parties would ever be able to keep up with fact checking the flood of posts on Meta’s platforms, so maybe this was necessary. But a real drawback to Community Notes is that they are slow, and posts with misinformation can be seen by millions of people before a note appears.
Moving the Trust & Safety team to Texas
In a curious move, Zuckerberg says the Trust & Safety team will be moved from California to Texas to “help remove the concern that biased employees are overly censoring content.”This seems based on vibes, rather than any actual evidence that California teams “censor” more than Texas teams. And it’s insulting to both Californians and Texans.
Bring back “civic content”
Meta currently requires opting in to be recommended “political” content from accounts you aren’t following, with a broad definition of “political”. They will be phasing this content back in, ”while working to keep the communities friendly and positive.”The political content recommendations will be based on personalized signals from the posts you view and engage with. They will also expand the controls for how much political content you will see.
I think Meta’s definition of political content was always overly broad, not only including “government” and “elections”, but also “social topics that affect many people”. Social topics could include politicized issues like gender, immigration, climate change, and health policies.
But I’m not sure how you can have both friendly and positive communities, along with more recommendations of political social topics and less moderation of hateful content.
Pushing back against “Foreign Governments”
Zuckerburg wrote that Meta will “work with President Trump to push back against foreign governments going after American companies to censor more.”While he did not provide any specifics, this could include the EU’s Digital Services act that requires very large platforms to lessen the impact of misinformation.
How will this change Meta’s platforms?
I would love to see the tech press push Zuckerberg to say the quiet words out loud. What kind of content does he see California teams “overly censoring”? What exactly are the foreign governments asking Meta to moderate? Is there any plan to try to stop the spread of misinformation?But I expect most news sites will republish Meta’s statements without pushing for more details.
Will there be a flood of posts promoting conspiracy theories and hateful content? Probably. But I think it’s unlikely to get as bad as X, where site owner Elon Musk is also the most followed account, and is actively promoting such content.
What will they do when the Trump administration asks for content to be removed? This will happen, and the response will show whether Zuckerberg is actually interested in free expression.
I expect Zuck will do whatever he thinks is best for Meta’s business, and he doesn’t care much about how that affects actual people.
I expect Zuck will do whatever he thinks is best for Meta’s business, and he doesn’t care much about how that affects actual people.
What people are saying
X CEO Linda Yaccarino on Meta’s shift to Community Notes: “Mark, Meta - Welcome to the party.”Mark Zuckerberg says people leaving Meta's platforms after this change are just "virtue signalling".
John Herrman at New York Magazine's Intelligencer, writes about "Mark Zuckerberg's eternal apology tour."
Andrew Hutchinson @ Social Media Today: Will Meta's revised approach to moderation impact its ad business? His conclusion: maybe, but the broad reach of brands on Facebook and Instagram might make them reluctant to make a moral stand.
Civil rights lawyer and former president of the NAACP Sherrilyn Ifill notes that these aren’t new opinions for Zuckerberg.
Disinformation researcher and Cofounder of the University of Washington Center for an Informed Public Kate Starbird notes that fact checkers actually took extra cautions to not appear biased, and there is no data to support anti-conservative bias in fact checking. Conservatives just share more false information.
GLAAD’s response notes “Today’s sweeping and extreme policy changes represent a wholesale abandonment of the norms and best practices of content moderation.”
Civil rights lawyer and former president of the NAACP Sherrilyn Ifill notes that these aren’t new opinions for Zuckerberg.
Disinformation researcher and Cofounder of the University of Washington Center for an Informed Public Kate Starbird notes that fact checkers actually took extra cautions to not appear biased, and there is no data to support anti-conservative bias in fact checking. Conservatives just share more false information.
GLAAD’s response notes “Today’s sweeping and extreme policy changes represent a wholesale abandonment of the norms and best practices of content moderation.”
Casey Newton @ Platformer: Meta surrenders to the right on speech: “I really think this a precursor for genocide,” a former employee tells Platformer.
Casey Newton @ Platformer: Inside Meta’s dehumanizing new speech policies for trans people
Tech Crunch: Mastodon CEO calls Meta’s moderation changes ‘deeply troubling,’ warns users cross-posting from Threads
Kate Knibbs @ Wired: Meta Now Lets Users Say Gay and Trans People Have ‘Mental Illness’
Mike Masnick @ TechDirt: Meta’s Moderation Modifications Mean Anti-LGBTQ Speech Is Welcome, While Pro-LGBTQ Speech Is Not
Clare Duffy @ CNN: Calling women ‘household objects’ now permitted on Facebook after Meta updated its guidelines
Clare Duffy @ CNN: Calling women ‘household objects’ now permitted on Facebook after Meta updated its guidelines
Electronic Frontier Foundation (EFF): Meta’s New Content Policy Will Harm Vulnerable Users. If It Really Valued Free Speech, It Would Make These Changes
Media Matters for America: Zuckerberg and Meta are done pretending to care about mitigating the harms their platforms cause
Media Matters for America: Zuckerberg and Meta are done pretending to care about mitigating the harms their platforms cause
Taylor Lorenz @ User Mag: Stop pretending Meta cares about free speech
404 Media reports: ‘It’s Total Chaos Internally at Meta Right Now’: Employees Protest Zuckerberg’s Anti LGBTQ Changes
Donald Trump says the changes were “probably” due to his threats against Meta and Zuckerberg.
Donald Trump says the changes were “probably” due to his threats against Meta and Zuckerberg.
My take is much more practical: I think Zuck has seen real user engagement plummeting on Meta platforms, and seen the popularity of Community Notes on X. He's going where the engagement (aka $$$) is. He's seen Community Notes and the often snarky way people use them as just another engagement metric rather than the "community-based fact checking or contextualizing" they pretend to be. Nobody likes the wet blanket of a 3rd party fact checker, but getting to see someone from "your side" community note somebody from the "other side" draws clicks, views, engagement. Just my opinion on this.
ReplyDeleteMaybe? It's hard to know what Zuck is thinking. Some of these changes probably save Meta money too. But I feel like some of it is resentfulness at those wet blanket fact checkers suggesting some content should be moderated.
Delete