Even as news organisations and fact checking outlets pushed back on Prime Minister Narendra Modi and the Bharatiya Janata Party’s claims that the Congress’s manifesto promised to divert resources from Other Backward Classes and Scheduled Castes to Muslims, one fact checking mechanism notably failed: Community Notes on X.
The feature was launched in India on April 4, and allows users to flag misleading posts by adding additional context, with users voting for whether other users’ ‘notes’ are useful or not.
Also Read | BJP video removed from Instagram after ‘hate speech’ complaints
No posts on X by PM Modi or the BJP on the subject have been successfully flagged by the feature, a review by The Hindu found. The feature appears to not be displaying fact checking notes at all for polarising content from the BJP over the past few weeks. That does not mean that people are not writing draft community notes on such content — rather, notes in total disagreement to any given post are being submitted but these community notes are not actually being approved and shown to X users.
Users who have been approved to suggest Notes on X have flagged some posts by PM Modi featuring the rhetoric that he and his party have pushed in recent days. One video posted on May 7 features Mr. Modi urging voters to “choose between Ram Rajya and Vote Jihad”. One note submitted by a user under this post said Mr. Modi appeared “to target groups, inciting divisions,” and that doing so was a model code of conduct violation, while another argued that since the term was coined by a Samajwadi Party leader (something Mr. Modi does not mention in the video), he was actually “calling the opposition out for [a] communal electoral conspiracy.”
A key problem may be the automation that Community Notes relies on: for a note to be visible, distinct groups of users who have disagreed in the past on other content must have consensus on a note. In a polarised political environment, Community Notes suggestions often show dueling points of view soon after an initial note has been suggested.
Why Community Notes didn’t work
Former Twitter executives in India and in the U.S. told The Hindu that X rolling out the Community Notes feature in India just weeks before the election without any human moderators was expected to have major flaws and could be perceived as an action to simply boost their brand.
“Content moderation is quite clearly against [the firm’s owner, billionaire] Elon Musk’s political philosophy and fiscal plans for the company and this feature Community Notes allows the company to claim that it is doing something to counter false narratives while not actually removing content or investing at the levels it did before regarding human trust and safety personnel,” said a former senior Twitter official in the U.S. who used to head up the platforms news, government and elections team. Mr. Musk, who acquired X in 2022, has largely pushed out trust and safety executives from the firm, and is a vocal proponent of no-holds-barred freedom of expression.
“To make Community Notes work X has to want to make it work, I think it’s pretty clear that’s not really a priority for X. The company just wants to check a box, so that’s exactly what it’s doing,” the former senior Twitter official in the U.S. said.
Kalim Ahmed, a disinformation researcher who is a registered Community Notes contributor (only those who are registered can view suggested notes; final notes are visible to all users), said that there were “clear attempts at manipulating” notes that were being written after Mr. Modi made the first of his controversial remarks in Rajasthan.
“If Community Notes is meant to make two people who disagree on politics or the election agree on something like facts and context, that’s not likely to happen,” a former Twitter India executive said (the firm had not been renamed when the executive departed). The executive, who declined to be named, said that “if [X] don’t have enough human moderators they would have also expected it to go a certain way which is what we’re seeing now.”
“But if it’s on another subject that’s less polarising and divisive than politics then Community Notes could be helpful and you could find the necessary agreement that the Notes rules have right now,” the former Twitter India executive said.
The executive and another former Twitter employee told The Hindu that Community Notes should strive to be more like Wikipedia, the free online encyclopaedia, edited by a community of anonymous volunteers called Wikipedians.
Wikipedia provides the closest analogy to a community moderated sources of truth or accuracy around sensitive topics that anyone can provide inputs or edits on but ultimately trusted Wikipedia human editors can vote the changes up or down, a sharp contrast to Community Notes’ complicated and automatised rating system.
What are Community Notes
Community Notes, which was first introduced in the U.S. in 2021, allows users to add informative notes to posts that might contain misleading information. Other users can mark these notes as “helpful” or “not helpful,” and based on this feedback, X highlights the most useful notes under the relevant posts.
Any user can leave notes on any post, and if enough contributors from different points of view rate that note as helpful, the note will be algorithmically approved and publicly displayed on the post. That note will be shown below the posts to all the users.
The platform has struggled to combat misinformation through the Notes feature for years in other countries.
X did not respond to a request for comment. The firm’s office in the national capital was visited by Delhi Police after the company’s own executives had flagged posts by BJP members as “manipulated media”. Simultaneously, as debates raged in 2021 onwards around whether social media companies should be “arbiters of truth,” Community Notes emerged as an alternative that let users themselves decide what content needed context.
Mr. Ahmed, the researcher, said that it was not clear that the attempts at suppressing Notes on political posts will succeed in the long term, as the feature has just been launched. Moritz Pilarski, Kirill Solovev and Nicolas Pröllochs, researchers at the University of Giessen in Germany who studied Community Notes, said in a paper studying the feature that “different fact-checking approaches complement each other and may help social media providers to optimise strategies to combat misinformation on their platforms.”
