Last week, Meta chief executive Mark Zuckerberg announced a major change in how content in the Meta stable (including Facebook, Instagram and Threads) will be fact-checked. Bottom line: It won’t be. At least not by Meta or its third-party agents.
Instead, Meta is looking approvingly at the Community Notes feature on Elon Musk’s X—essentially to get us, the users, to “fact-check” each other at very little cost to the platforms. Mr. Zuckerberg, in announcing the move, said that the moderating systems Meta has been using for the past decade “make too many mistakes” and result in “too much censorship.”
Mr. Zuckerberg’s announcement has been met with condemnation and alarm, but The Atlantic’s Ian Bogost points out (correctly, in my view) that “fact-checking” in the rigorous journalistic sense is difficult to apply to social media. As he writes, “you can’t nitpick every post from every random person, every hobby website, every brand, school, restaurant, militia lunatic, aunt or dogwalker as if they were all the same.” He sums up the end of Facebook’s fact-checking attempt with “good riddance.”
We all know there are serious problems with social media content—the lies, the sexism, the racism, the homophobia, the fraud, the extremism. This is not surprising, given that most of humanity is now on social media. Sites like Facebook reflect who we are: Some of us like cat videos and others are drawn to harmful, even illegal, content.
So what will content moderation now look like for Meta (and what does it look like for X)? It is important to note that Meta will continue to moderate certain content. Its new chief global affairs officer, Joel Kaplan, says the company itself will still moderate content related to drugs, terrorism, child exploitation, frauds and scams. For the rest, it is hoping to follow X’s Community Notes model.
Mr. Musk fired much of Twitter’s staff when he bought the social media platform and radically downsized its moderation and monitoring teams, essentially replacing them with its Community Notes feature, which X describes as aiming “to create a better informed world, by empowering people on X to collaboratively add helpful notes to posts that might be misleading.”
Community Notes contributors are X users who have signed up (and been accepted by X) to write and rate notes. Importantly, Community Notes does not operate according to “majority rules.” According to X, Community Notes are posted only if there is “agreement between contributors who have sometimes disagreed in their past ratings.” This is meant “to prevent one-sided ratings.”
X’s thinking is “if people who typically disagree in their ratings agree that a given note is helpful, it’s probably a good indicator that the note is helpful to people from different points of view.” One of the good things about Community Notes is that its underlying algorithm is public. However, it has been described as being “insanely complicated” by a technology writer at Yahoo!’s news site.
Moderation by ‘we the people’
The idea of moderating content by “we the people” is not new. Perhaps the most well-known experiment of this kind is Wikipedia, whose tagline is “the free encyclopedia that anyone can edit.” To be clear, Wikipedia is tricky. University students soon learn that Wikipedia, on its own, is not an acceptable source. Then again, it is widely recognized as a useful jumping-off point from which to research a topic.
But how well does X’s Community Notes program work? Yang Gao, an assistant professor at Gies College of Business at the University of Illinois Urbana-Champaign, led a team of researchers whose results showed that “receiving a displayed community note increases the likelihood of tweet retraction, thus underscoring the promise of crowdchecking.” In another paper, a group of researchers from the University of Luxembourg found that “exposing users to community notes reduced the spread of misleading posts by, on average, 61.4%.”
That all sounds positive and encouraging, but… (And there are a lot of buts.)
Who is able to post Community Notes? Theoretically, anyone with an X account that has been active for more than six months, who has a verified phone number, and who has no recent violations of X’s rules can participate in Community Notes. But the process of getting accredited by X is opaque, and there is no time limit for the process of accreditation. In November 2023, Yahoo! reported estimates by Mediawise, an organization run by the Poynter Institute,that there were only 133,000 Community Note users—and that while over 150,000 notes have been filed, only 10,000 or so have been posted so that they can be seen by X users. These numbers seem extremely low, given that X reportedly has over 500 million users globally. In my view, it is also a problem that the actual numbers of posted Community Notes and accredited Community Notes users are not publicly accessible.
In its announcement, Meta said it will be phasing in Community Notes first in the United States over the next two months. And this flags other big buts: What about the rest of the world? What about languages other than English?
The Community Notes program is still in its infancy, but the research on its effectiveness is promising. However, there is a danger of the rest of the world being left behind, without access to Community Notes and the ability to counter misinformation in languages other than English. It is difficult to discover exactly how many countries have X accredited Community Notes users. X says, coyly: “As there are important nuances in each market, we’ll expand the contributor base country-by-country,” but nowhere does it publish a list of countries which have accredited contributors.
Last September, the South African website Hypertext reported that Community Notes was available in only 44 countries worldwide—out of 193 members of the United Nations. While this figure might now be out of date, it is clear that not all countries in which X is available have the same access to Community Notes.
Africa, for example, is a continent of 54 countries and over 1.5 billion people, but until a few months ago, it did not have a single accredited Community Note user. This has now changed, with South Africa (where I live) being the first African country in which X has invited users to register for Community Notes. I registered but have yet to hear from X, and I have yet to see any Community Notes on South African or broader African issues on X.
Both X and Meta are clearly prioritizing Americans and those who speak English. This isn’t going to be good enough. The political and social ravages of misinformation are a global scourge. Why should those of us who use X and Meta be largely unable to benefit from their strategies to counter disinformation only because we live in Asia, South America, Oceania or Africa?
Doubtless the previous fact-checking/moderation practices at both Meta and X were problematic and in need of an overhaul. And the Community Notes alternative holds promise. But if Community Notes—moderation by platform users with differing political and social views—is indeed to be the alternative to formal third-party fact-checking, then everyone using the platform has to be able to participate in Community Notes. That is, people from every country, and speaking every language that a particular social media platform is available in, have to be able to participate in the broader Community Notes project to counter misinformation on that platform.
[For another perspective on Meta’s move away from fact-checking, check out: Facebook, fact-checking and the case for giving legacy media another chance]
