Meta’s decision to scrap fact-checking on Facebook and Instagram is under review after Brazilian football legend Ronaldo Nazario was caught up in a deepfake scam.
The company’s oversight board has launched an investigation after Meta initially rejected calls to remove a post that used an AI-manipulated video of Ronaldo to promote a gambling app.
Meta has now reversed that decision, admitting that it violated its rules on content attempting to use “a fake persona or to pretend to be a famous person in an attempt to scam or defraud”.
But its oversight board, which handles escalated complaints relating to Facebook, Instagram and Threads, says the case merits closer scrutiny of Meta’s policy on scams.
Part of that will see it examine how the social media giant’s decision to end third-party fact-checking last month “could impact the amount of deepfake endorsements”.
Meta CEO Mark Zuckerberg announced the controversial move on 7 January, saying its platforms would instead adopt a “community notes” system like that in use on X.
Zuckerberg said that some fact-checking was “biased” and that the policy change would allow the company “to get back to our roots around free expression”.
It came ahead of the second presidential term of Donald Trump, some of whose supporters have argued that online moderation amounts to censorship.
But critics of the move have questioned the efficacy of the community notes model and claim that it could lead to a rise in false and misleading information.
Meta watchdog seeks public comments on Ronaldo Facebook scam
The Ronaldo scam, which appeared in September last year, encouraged users to download the Plinco app, for playing a game that sees a ball drop down a peg-filled board.
Using video of former Real Madrid, Barcelona, Inter Milan and AC Milan striker Ronaldo and audio imitating his voice, it claimed players could earn more than their salary.
Clicking the link, however, took users not to the Plinco app but to a different game called Bubble Shooter. The post was viewed around 600,000 times.
The oversight board can recommend that Meta changes its policies. As part of its investigation, it has launched an appeal for public comments on three aspects of the case:
“The socioeconomic impact of endorsements by deepfakes imitating public figures on the public and the figure being imitated, especially in Brazil.
“The effectiveness of Meta’s enforcement practices for its policies against scams, specifically for content that contains fake personas and the impersonation of public figures, in Brazil and other regions.
“How Meta’s announcement on January 7, 2025, about ending proactive enforcement of certain policies could impact the amount of deepfake endorsements on Meta’s platforms, particularly in regions where user reports are less frequent.”