How do context notes work?
Contextual notes are a tool for crowd-sourcing content. They appear below potentially misleading posts.
Twitter has been using them since 2021 and in 2022 they became widespread on the social network, bought by Elon Musk in 2022 and renamed X.
Notes are proposed and written by volunteer users, who must register in advance, and are not edited by the X teams.
“Other users will then evaluate whether the article is useful or not, based on different criteria, such as the relevance of the sources and the clarity of the information,” Lionel Kaplan, president of the Dicenda social media content creation agency, told AFP.
“If there are enough positive ratings for the story, it will appear below the tweet to provide additional information,” he said.
The ratings “take into account not only the number of contributors who rated a post as helpful or unhelpful, but also whether the people who rated it appear to come from diverse backgrounds,” X explains on its website.
The principle is the same as Wikipedia. “We rely on the most active users of a social network or platform to increase the quality of content,” Kaplan added.
Meta, which announced that its grading program will be similar to X’s, considers it a “less biased” system than fact-checking.
What are the risks?
Facebook has a fact-checking program in more than 26 languages that pays more than 80 media outlets around the world to use their “fact-checks” on its platform, on WhatsApp and on Instagram.
“The problem with ratings is that verification depends on the crowd,” says Christine Balagué, a professor at the Mines-Télécom Institute and founder of the “Good in Tech” research network, which works on disinformation.
“The crowd can say the right thing, but there can also be malicious people there who are there to spread misinformation,” he says.
“This decision will affect users who want accurate and reliable information,” said Angie Drobnic Holan, US head of the International Fact-Checking Network (IFCN) in X.
“Fact-checkers have never shown bias in their work and these criticisms come from people who think they can exaggerate the facts and lie without being refuted or contradicted,” he added, highlighting the climate of political pressure in the United States on the eve of the inauguration of President-elect Donald Trump.
Bill Adair, co-founder of the IFCN network, said it was “disturbing to see Mark Zuckerberg echoing political attacks on fact-checkers because he knows that those who participated in his program were signatories to a charter of principles that calls for transparency and impartiality.”
Trump, who has been particularly critical of Meta and its boss in recent years, accusing the company of liberal bias, said that “probably” influenced the decision.
Are ratings effective against misinformation?
Researchers have shown that context notes “reduce the spread of misinformation by around 20%” on the X network, admitted Christine Balagué. And although this system “is not 100% reliable,” Lionel Kaplan believes that context notes are effective.
He believes that they allow for the processing of a larger volume of information than fact-checking, which is “cumbersome to implement.” He also points to a “democratic” functioning, which allows “having and confronting different points of view.” X, however, is regularly accused of allowing misinformation to spread since Elon Musk drastically reduced the moderation teams.
But, according to this expert, this is mainly due to the fact that “X encourages radicalism” and, consequently, makes false information more visible than in other places.
Meta said users could begin signing up as early as Monday to write notes when the program launches, without giving details on eligibility criteria.
The company also announced it will give users more control over how much political content they want to see on Facebook, Instagram or Threads.
It remains to be seen whether this system will increase the spread of false information. According to Kaplan, Meta is not necessarily interested in highlighting controversial content as it might upset advertisers.
“Meta must comply with regulations on content moderation”
The European Commission took “note” on Tuesday of the decision by the multinational Meta to end its data verification program with third parties in the US, and warned it that the European Union is governed by the Digital Services Act (DSA) and that it is obliged to comply with the obligations it sets out.
“We have taken note of Meta’s announcement regarding its content moderation practices in the US. We have no comment on this matter. The Digital Services Act applies in the EU,” European Commission spokesperson for technological sovereignty Thomas Regnier told EFE.
Meta CEO Mark Zuckerberg announced in a statement today that the tech giant, which runs platforms like Facebook, Instagram and WhatsApp, will do away with the accredited fact-checkers it has been using in the United States and replace them with a user-based rating system similar to that of the X network. A change that, for the moment, will only take effect in the United States.
“We continue to monitor compliance of VLOPs (the largest online platforms) with their DSA obligations in the EU,” the spokesperson for the European Executive stressed.
Under the DSA, “collaborating with independent fact-checkers can be an effective way for platforms to mitigate systemic risks arising from their services, while fully respecting freedom of expression,” Regnier said.
And he recalled that this applies to risks such as the spread of misinformation or the negative effects on civic discourse and electoral integrity.
Under EU law, “before implementing functionalities that may have a critical impact on systemic risks, large online platforms must conduct a risk assessment and submit a report to the Commission services,” the spokesperson said.
She also noted that Meta is a signatory to the Code of Practice on Disinformation, which contains several specific commitments on fact-checking.
