X Community Debates: Is Fighting Fake News Worth Restricting Free Speech?

Elon Musk’s endorsed moderation tool, Community Notes, which was intended to empower Twitter users in monitoring the platform for misinformation, has faced criticism from experts and politicians due to its perceived abusive corrections. The tool has now been renamed X.

What are Community Notes?

Announced in early 2021 — two years before Musk bought the platform — as Birdwatch, the feature initially allowed volunteers in the United States to report questionable messages and provide context for notes.

These notes were initially only visible on a separate site, but then below the message itself.

Twitter had just launched Donald Trump from a platform that incited violent riots at the US Capitol and sought to promote healthier conversations.

In November 2022, a community memo prompted the White House to retract a tweet that had exaggerated the impact of President Joe Biden’s retirement benefits.

Upon becoming the platform’s new owner, Musk played down the moderation, but still said his “goal is to make Twitter the most accurate source of information on Earth, regardless of political affiliation.”

The publication of Community Notes has since been implemented in 44 countries.

How does it work?

Twitter users registered as contributors write suggestions as notes and are not edited by X employees. However, they must follow X’s moderation rules.

Suggested notes are then delivered to other participants. Those that get enough votes to be useful can then be selected by an algorithm for publication.

The algorithm is inspired by the one Netflix uses to recommend content, and aims to “identify notes that appeal broadly across perspectives,” rather than just those that get the most upvotes.

It is updated regularly, and was recently announced to limit the replacement of notes after publication.

Is it effective in combating disinformation?

According to X, people are on average 30% less likely to agree with a post’s content after reading the attached community posts, and they’re also less likely to repost it to their followers.

But Alex Mahadevan, program director of the Poytner Institute’s MediaWise digital literacy initiative, pointed out that the flaw in the algorithm is that it requires consensus across ideological divides to publish posts.

“Maybe it would have worked four years ago,” he said at a conference in South Korea in June.

“It doesn’t work anymore because 100 people on the left and 100 people on the right don’t agree that vaccines are effective.”

What was required, he explained, was an “ideological agreement on the truth,” and in an increasingly partisan environment, he said, achieving that consensus is nearly impossible.

Some hold Musk to be partly responsible for this environment, as since taking over X he has significantly cut content policing teams and pulled the platform out of the EU’s voluntary disinformation best practices.

For Musk, community notes are an economical way to moderate content, including political advertising, which has recently been re-accepted on the platform.

For Mahadevan, the system has proven effective at identifying and adding context to non-political content, such as pointing out AI-generated images, misleading advertising and conspiracy theories that have already been debunked.

Is the algorithm neutral?

Although the algorithm’s complex mathematical formula is intended to prevent manipulation, it is not foolproof and requires a large number of evaluators.

Julien Pain, host of Franceinfo’s True or Fake, said there are groups in France that game the algorithm to promote their ideas.

He said hard-right groups flood the forum with notes from left-wing politicians and try to get notes from conservative posts removed.

“Donations are not always accepted, but some do,” he told AFP.

“A comment is interesting when it adds a bit of factual context to nuance the statement. But these are just declaring it false and trying to destabilize the person,” he added.

For Julien Bayou, a French Green lawmaker, the notes have become “a tool for ideological competition that, unfortunately, can become distorted.”