Elon Musk Might Not Trust Fact-Checkers, But X Users Clearly Do

Carlos Hernández-Echevarría is Assistant Director, Head of Public Policy & Institutional Development at Fundación Maldita.es. This piece references a new report from Maldita.es, Faster, trusted, and more useful: The Impact of Fact-Checkers in X Community Notes.

Elon Musk has announced he is “working to fix” Community Notes after X users dared to upvote and make visible some notes that he disagrees with. That is a stark departure from his “Vox Populi, Vox Dei” mantra (the voice of the people is the voice of God) and from his repeated assurances that he would not meddle with the decisions made by the people with “diverse viewpoints” that propose and rate the notes. But perhaps more importantly, it indicates that not all X users share Mr. Musk’s disdain for factual information.

Since he purchased Twitter (now X) in 2022, Musk hasn’t minced words when sharing his opinions about professional fact-checkers. “Liars” is one characterization he prefers when not directly referring to them as “evil.” His professed distrust for all we do appears to be as strong as his faith in the potential of X’s Community Notes to replace the need for fact-checkers, a faith that Meta founder and CEO Mark Zuckerberg has also recently adopted.

But the reality of the relationship between fact-checking organizations and Community Notes is more complicated. At Fundación Maldita.es we analyzed almost 1.2 million Community Notes proposed by X users last year, whether made visible on the platform or not—all that were available for download. We found that professional fact-checking organizations are the third most referenced source on Community Notes worldwide right after X itself and Wikipedia (also a recent target of Musk’s ire).

In all, 1 in every 27 Community Notes proposed globally listed a link to a fact-checking organization, which suggests X users have an altogether different appreciation for professional fact-checkers than the owner of the platform. Moreover, the data from visible Community Notes also allows us to appreciate how deep that trust is.

Faster and more useful

Both Musk and X have stressed how key it is for a Community Note to have the backing of “people from a wide range of viewpoints” before it can appear alongside a problematic tweet. That is why few notes—only 8.3%—are ever rated as useful by enough “diverse” users to become visible. However, a note using a fact-checking article as a source is almost twice as likely to generate that kind of trust—the chance a note will cross the threshold required to surface visibly on the platform increases to 15.2% when it cites an article by a European fact-checker.

The X Community Notes data also contains interesting evidence that counters the commonly held view that professional fact-checkers are slower at addressing misinformation than crowdsourcing. Notes that cite fact-checks take less time than others to be proposed after the original post (a median of 4 hours and 25 minutes, 23 minutes less than a typical note), and they are rated as useful and become visible alongside the post much sooner (5 hours and 40 minutes, 24 less than the median time for all notes). All in all, they are visible 90 minutes earlier than all notes.

As much as these data points are a vindication of the work of fact-checkers on a platform whose owner and loudest voice relentlessly attacks them, they should not be understood as an endorsement of X’s current approach to Community Notes. It’s good news that professional fact-checking does better than other sources in trust, speed, and visibility, but a focus on “consensus by the different” should never take precedence over “factuality” when it comes to addressing harmful misinformation at scale.

Despite facing frequent mischaracterization by powerful figures, fact-checkers in Community Notes have proven their strength. Even in the toughest conditions, our work remains valued by the public for its factual, nonpartisan integrity. Perhaps Elon Musk—and Mark Zuckerberg—should look at the data.