SAN FRANCISCO: Independent monitors voiced concern Tuesday that Meta’s recent decision to ditch fact checks on their Facebook platform could endanger human rights.
Meta’s surprise announcement in January that it was ending its US fact-checking program triggered scathing criticism from disinformation researchers who warned it risked opening the floodgates for false narratives.
Now, the board that serves as the top court for Meta content moderation decision says the social media giant’s announcement about policy and enforcement changes to the handling of hateful and potentially harmful posts was made “hastily,” according to a statement by the Meta Oversight Board released Tuesday.
“People have the right to express controversial opinions,” said board co-chair Helle Thorning-Schmidt.
“People should also be safe from harm.”
Meta Chief Executive Mark Zuckerberg released the news in a sweeping policy shift that analysts saw as an attempt to appease then US President-elect Donald Trump, who has equated checking facts with censorship.
As Meta rolls out the moderation changes globally, it is essential that the tech giant mitigate human rights perils that may result from a reduction or absence of fact-checking, according to the board.
The oversight board made 17 recommendations, including that Meta assess the effectiveness of Community Notes compared to third-party fact-checking, “particularly in situations where the rapid spread of false information creates risks to public safety.”
Meta had employed third-party fact checkers, AFP among them, to expose misinformation disseminated by the platform.
Zuckerberg said Meta’s platforms, Facebook and Instagram, would instead use “Community Notes similar to X” in the United States to police accuracy of posts.
Community Notes is a crowd-sourced moderation tool that X, formerly Twitter, has promoted as the way for users to add context to posts, but researchers have repeatedly questioned its effectiveness in combating falsehoods.
“You wouldn’t rely on just anyone to stop your toilet from leaking, but Meta now seeks to rely on just anyone to stop misinformation from spreading on their platforms,” Michael Wagner, from the School of Journalism and Mass Communication at the University of Wisconsin-Madison, told AFP when Meta announced the change.
“Asking people, pro bono, to police the false claims that get posted on Meta’s multi-billion dollar social media platforms is an abdication of social responsibility.”
While Meta has vowed to honor the board’s rulings on appeals of its decisions to take down or leave up posts, the tech company does not have to abide by its policy recommendations.

LONDON: Italy’s far-right League party has been referred to the country’s communications watchdog after opposition parties filed a complaint over “racist, Islamophobic and xenophobic” images generated by artificial intelligence and shared on social media by deputy prime minister and party leader Matteo Salvini.

The complaint was submitted to Agcom, Italy’s communications regulatory authority, on Thursday by the center-left Democratic Party, along with the Greens and Left Alliance. It alleges the images published by the League contained “almost all categories of hate speech,” according to The Guardian, which first reported the story.

“In the images published by Salvini’s party and generated by AI there are almost all categories of hate speech, from racism and xenophobia to Islamophobia. They are using AI to target specific categories of people — immigrants, Arabs — who are portrayed as potential criminals, thieves and rapists,” said Antonio Nicita, a PD senator.

Nicita also criticized the decision to blur the faces of the supposed victims, calling it “deceptive” and accusing the League of intentionally misleading users into believing the images were real.

Emilio Borrelli, an MP with the Greens and Left Alliance, said the images were “part of their strategy to create fear among citizens” and “incite hate.”

Over the past month, dozens of apparently AI-generated images have been posted across the League’s social media channels, including Facebook, Instagram and X. Many depict men of colour, often armed with knives, attacking women or police officers.

A spokesperson for Salvini’s party confirmed some of the pictures were digitally generated but insisted: “The point is not the image. The point is the fact,” adding the posts were “based on true reports from Italian newspapers.”

However, AI forensic experts have stated all the images in question bore clear signs of being artificially generated. They also noted that while platforms are required to label AI-generated content, in most cases automatic detection tools failed to do so.

In one of the posts cited in the complaint, a mother and father in Islamic dress appear to be shouting angrily at a young girl — a portrayal the complainants say fuels racial and Islamophobic stereotypes. The newspaper cited in the post, Il Giorno, makes no reference to the family’s religion and does not include any photographs. The only detail given was that the child had attended Arabic language classes.

As The Guardian reported, the use of AI-generated imagery by far-right parties across Europe has surged in recent months. The targets are often refugees from conflict zones such as Syria, Sudan and sub-Saharan Africa, as well as people from other minority backgrounds. These depictions frequently invoke the debunked “Great Replacement” conspiracy theory, which falsely claims that immigration is part of a plot to erode European identity and culture.

Salvini, who has capitalized on rising refugee arrivals in Europe to maintain a prominent role in Italian politics and advocate for stricter immigration policies, has frequently made headlines for inflammatory remarks, including calling immigrants — often men — “dogs and pigs.” In late 2024, he was acquitted of charges of kidnapping and dereliction of duty after judges ruled that the evidence presented by prosecutors was insufficient to convict him. The case stemmed from a 2019 incident in which Salvini, then interior minister, refused to allow a Spanish migrant rescue ship to dock in an Italian port, leaving those on board stranded at sea for 19 days.

Asked whether the League was aware the images could incite hate, a party spokesperson said: “We are sorry, but our solidarity goes to the victims, not the perpetrators. If denouncing crimes committed by foreigners means ‘xenophobia’, perhaps the problem is not the word but those who use it to censor debate. We will continue to denounce, with strong words and images, what others prefer to ignore.’’

If Agcom finds the League’s content in violation of regulations, it could act under the EU’s Digital Services Act, which allows it to order the removal of posts, shut down accounts or impose fines on social media platforms for failing to moderate harmful content.