After Years of Moderation-Heavy Zagging, Zuckerberg Announces That Meta Is Going Back to Moderation-Light Zigging Across Its Platforms

Tuesday, 7 January 2025

Mark Zuckerberg today announced major changes to the way Meta is going to apply content moderation across Facebook, Instagram, and (I presume, Threads). His main announcement is a video, for which there’s an unofficial transcript here. Zuckerberg himself summarized his own points in a thread on Threads, which I’ll annotate below:

1/ Replace fact-checkers with Community Notes, starting in the US.

This sounds ... fine? Just about everything at Twitter/X has taken a turn for the worse under Musk, but community notes seem to be a model that actually works pretty well. My favorite two social platforms right now, by far, are Bluesky and Mastodon. Neither has fact checkers. And I’ve never seen any evidence that Meta’s fact checkers have ever checked a single fact. (Maybe that’s what the team is doing in Macrodata Refinement at Lumon?)

2/ Simplify our content policies and remove restrictions on topics like immigration and gender that are out of touch with mainstream discourse.

This too seems fine. Let people spout what they want to spout. The problem has never been about whether users should be allowed to express opinion A about topic Z. The problem has always been about which opinions algorithmic platforms choose to promote. I can frame this specifically about Twitter/X. Under Elon Musk’s ownership, you can express pretty much whatever political opinions you want. That’s great. But, also, in my experience during the Musk era — and I’m not sure anyone would disagree with this — the political opinions that tend to get algorithmically boosted are those that align with the Musk/Trump agenda.

I don’t want to make a hypothetical example. I went to X.com just now, and searched for “Hillary Clinton medal of freedom”. In reverse order, the third tweet in my search results (screenshot) was a congratulatory tweet from the Clinton Foundation account. The second was a tweet from Hillary Clinton’s own account, thanking President Biden, accompanied by a photo of Biden fastening the medal around her neck at the White House. But the first tweet in my results was from “Dr. Clown, PhD”, presenting an AI-generated video of Clinton transmogrifying into a fire-spewing horned demon.

I legit think it’s fine that the esteemed Dr. Clown is permitted to post such a video to X. It doesn’t need to be fact-checked. (Although someone should let Tucker Carlson know it’s not real.) It certainly shouldn’t be forbidden. But I also have zero interest in searching for political news on a platform that would promote such un-clever AI-generated slop when searching for “Hillary Clinton medal of freedom”, let alone promote it to the very top spot in the results, literally above Hillary Clinton’s own tweet about the award.¹

The question isn’t what sort of posts Meta is now going to allow, but rather, what sort of posts are their algorithms going to promote, and to whom. I’m not trying to be a Pollyanna here — I’m fully aware that the sort of people who think that Twitter/X has improved, not declined, under Elon Musk’s ownership are the sort of people who not only think their posts should be permitted, but that they should be promoted to a wide audience, and that anything less than wide algorithmic promotion is the result of nefarious “shadowbanning”. If Meta intends to start showing me content from the likes of the esteemed Dr. Clown, well, I’m out. If they simply want to show more Hillary-Clinton-is-a-demon AI slop to those who think that’s clever, that fine.

So, let’s see what Meta means here. In a weird way I sort of hope this means Meta’s platforms go full virulent Nazi. I’d love to pay as little attention to another social media platform as I now do to X.

3/ Change how we enforce our policies to remove the vast majority of censorship mistakes by focusing our filters on tackling illegal and high-severity violations and requiring higher confidence for our filters to take action.

Meta would do well to illustrate this policy change by citing specific examples of content that they have censored, but no longer will. Without specific examples this is meaningless, and could result in anything from Meta’s platforms changing dramatically to not changing at all. Admitting you’ve made mistakes without citing what those mistakes were isn’t really admitting to anything other than a change in vibes. (This largely applies to the entirety of Zuckerberg’s announcements today.)

In an accompanying post on the Meta company blog (still hosted, oddly, at the about.fb.com domain), newly-promoted chief global affairs officer Joel Kaplan — a career Republican who served as an administration official for all 8 years of George W. Bush’s two terms — writes:

Over time, we have developed complex systems to manage content on our platforms, which are increasingly complicated for us to enforce. As a result, we have been over-enforcing our rules, limiting legitimate political debate and censoring too much trivial content and subjecting too many people to frustrating enforcement actions.

For example, in December 2024, we removed millions of pieces of content every day. While these actions account for less than 1% of content produced every day, we think one to two out of every 10 of these actions may have been mistakes (i.e., the content may not have actually violated our policies).

With millions of examples and, supposedly, a 20 percent false positive rate, there should be plenty of examples they could cite. But Kaplan cites not a one. It could be empty posturing. It could be a lurch to promote outright right-wing viewpoints. Only their actual actions will reveal the answer.

Back to Zuckerberg:

4/ Bring back civic content. We’re getting feedback that people want to see this content again, so we’ll phase it back into Facebook, Instagram and Threads while working to keep the communities friendly and positive.

On the surface this seems to translate as, roughly, “We’ve downplayed political content in recent years, most conspicuously in 2024, an election year, and now that the election is over and Trump is coming back into office, sure, let’s bring political content back.” Which seems like the wrong timing, to me, but whatever. I don’t see this timing as Zuckerberg putting his thumb on the scale in favor of Trump. Meta has been de-prioritizing all political content, and if they’re really going to re-prioritize it now, that would apply as much to Trump’s critics as his supporters.

Consider if the timing had been reversed, and Meta’s platforms had been a political free-for-all throughout 2024, with political takes promoted to the top of everyone’s feeds, and today, two weeks before Trump takes office again, Zuckerberg announced that Meta was going to reduce the heat in the proverbial kitchen by de-emphasizing “civic content” across all its platforms. Those determined to view any decision he makes as being in the bag for Trump — which may or may not be correct — would be jumping up and down today arguing that Zuckerberg is shamelessly attempting to silence Trump’s critics on the cusp of his second inauguration.

What to me seems obvious is that in recent years, Meta made a decision to eschew controversial content even at the cost of engagement. How they think they can re-prioritize political (or in Zuck’s euphemistic phrasing, civic) content while “keeping the communities friendly and positive”, I don’t see. Nothing is going to be “friendly and positive” in our political discourse so long as Trump remains a national figure, and he hasn’t even re-taken office yet. What the whole thing primarily highlights is that Zuckerberg has no guiding principles — zero, zilch, nada — behind any of Meta’s platforms other than “success” in and of itself. Is it about you and your closest friends and family? Or about following celebrity influencers? Politics yes, or politics no? The answers were all different a few years ago, and they’re likely to be different again a few years from now.

5/ Move our trust and safety and content moderation teams out of California, and our US content review to Texas. This will help remove the concern that biased employees are overly censoring content.

First, California and Texas are roughly counterbalanced political opposites, but Texas is slightly less red than California is blue. Consider the recent election.

California:

• Kamala Harris: 9,276,179 votes (59%)
• Donald Trump: 6,081,697 votes (38%)

Texas:

• Donald Trump: 6,393,597 votes (56%)
• Kamala Harris: 4,835,250 votes (42%)

Texas is a big state, though. If Meta is moving its content moderation team to Austin, they’re moving to a county that voted for Kamala Harris 69–29, significantly more “blue” than California as a whole.

But this is all performative. Zuckerberg’s geographic explanation makes no sense so long as Meta itself is headquartered in Menlo Park, California. Are we to believe that any biases in Meta’s content moderation are determined by the moderators themselves — who are largely, if not entirely, contractors who work for third party companies? In 2019, Casey Newton wrote a spectacular piece at The Verge profiling several of the “800 or so” employees working as Facebook content moderators for a vendor named Cognizant at a site in Tampa, Florida. (Florida’s 2024 election results were nearly identical to Texas’s, 56–43 for Trump.)

It’s the algorithms that drive discourse on Meta’s platforms, and the engineers who steer them work in California. But really, whether the entire staff of the company works in Wyoming or Vermont doesn’t matter. Ultimately, everyone at Meta answers to Zuckerberg.

6/ Work with President Trump to push back against foreign governments going after American companies to censor more. The US has the strongest constitutional protections for free expression in the world and the best way to defend against the trend of government overreach on censorship is with the support of the US government.

Mark Cuban’s theory on how this could work, in a reply to Zuckerberg on Threads:

Translation: Americans are going to see Tariffs on products from countries you believe censor Meta services as a means of pressuring them into removing any restrictions that impact your profitability in those countries.

Also: You’ll have carte blanche to take posts that no longer have restrictions, making them a more explicit representation, and train your AI Models

Also: Bluesky is now the only moderated social media platform with tools that allows the user to control their own experience

To me, this just seems like Zuckerberg donning a MAGA cap and saying “Fuck yeah, Trump!” to please the boss. I don’t see the Trump administration altering its tariff plans for the EU based on the EU’s content moderation and data privacy policies for U.S. social media companies. And I certainly don’t see the EU changing its content and privacy regulations based on Trump-imposed tariffs on unrelated physical goods. But who knows? Nobody. Trying to predict what Trump might actually do is like trying to predict the shuffle of a fair deck of cards.

And ultimately, that’s where Zuckerberg and Meta seem most aligned with Trump. My take on Trump post-election has been to stop paying attention, as best I can, to anything he says. I’m only paying attention to what he does. With any other national leader, there’s a correlation between their words and their eventual actions that makes paying attention to what they say worthwhile. With Trump, there’s almost no correlation, and his endless stream of outrageous proclamations are nothing but a distraction.

The same goes for Meta. Remember their “Oversight Board”? It’s still there, supposedly. They even piped up today, with a pathetic “Hey Mark, remember us? We’re still here” statement. Turns out they’re a complete joke with no actual influence, let alone authority, over the company’s platforms. Shocker. But much of the news media fell for the idea that the Oversight Board was in any way a serious endeavor hook, line, and sinker over the years of their pantomime existence.

Today’s Meta announcements sound significant, but so too does the notion that the U.S. military might invade Greenland to claim its territory. Don’t get distracted by blather. Let’s see what happens.