For anyone who has watched a rumor mushroom into a family group chat argument, or seen a grainy video falsely labeled as current events race across a country’s timeline, the news out of Meta this week felt like a quiet shift in the tectonic plates of the internet. The company that owns Facebook, Instagram, and Threads announced it is testing a system called Community Notes in 16 Latin American countries, a move that would effectively replace professional fact-checking with crowdsourced user reviews. As of now, the rollout is still a trial, and for a little while longer, accredited newsrooms and specialized agencies will keep flagging false content on Meta’s platforms across the region. But the direction is unmistakable: after years of investing in third-party verification, Meta is betting that ordinary users, not trained journalists, should decide what is true and what is misleading online. The announcement, made on a Wednesday, has sent ripples far beyond Silicon Valley. Across Latin America, media experts, academics, and disinformation watchers are bracing for what they say could be a serious erosion of public conversation. This is not a niche policy tweak; it is a decision about who gets to shape reality for hundreds of millions of people.
The system itself sounds simple, even democratic. Community Notes allows users to add context to potentially misleading posts, and if enough people across a range of viewpoints agree, that note becomes visible to everyone. It is the same model Elon Musk installed at X, formerly Twitter, after he bought the platform and dismantled its trusted fact-checking operations. But experts in Latin America are warning that crowdsourced moderation is not the same as professional verification. Social media users, they point out, do not generally apply the methodical, ethical, and journalistic standards that trained fact-checkers bring to their work. Carlos Rodriguez, a professor at Colombia’s Universidad de La Sabana, put it plainly: ordinary users are not fact-checkers, and they should not be expected to behave like them. The concern is not that people are stupid; it is that misinformation is engineered to exploit emotion, speed, and existing prejudice. A user scrolling at 2 a.m. is not likely to pause, verify sources, and weigh evidence before adding a note. More troubling, platforms may become a battleground where competing partisan crowds vote on the truth, rather than a space where accuracy is established through evidence. Ramon Salaverria, a professor at the University of Navarra in Spain, warned that as polarization and disinformation spread, relying on consensus between users can devolve into worrying situations. When truth is determined by popularity, the loudest and most coordinated groups—not the most accurate—are likely to win. Alejandro Martin del Campo, a media expert at Mexico’s Tecnologico de Monterrey, said the consequences could be considerable, making it easier for misinformation to spread more quickly. That is not a distant hypothetical. In a region already battling false narratives around vaccines, elections, natural disasters, and migration, the removal of a professional safety net feels like a dangerous gamble.
The organizations doing this work are not faceless bureaucracies. Around 14 independent fact-checking agencies in Latin America are part of Meta’s global verification program, which began in 2016. One of them is AFP, with its fact-checking services operating in 26 languages. These agencies investigate viral claims, track manipulated images, interview sources, and publish corrections that social media platforms then use to label or demote false content. It is slow, unglamorous, and absolutely essential. The day after Meta’s announcement, a consortium of these organizations, called LatamChequea, issued a statement urging Mark Zuckerberg’s company to rethink its decision. Their message was direct: Community Notes are not enough in contexts of high disinformation. They asked Meta to complement, not replace, independent verification. It was a polite but firm plea from people who have spent years watching misinformation mutate, travel, and cause real harm. For them, the move away from professional fact-checking is not an abstract ideological debate. It is about whether a mother in a rural village will see a health warning before a false cure goes viral, or whether voters will receive a correction before a fabricated rumor about a candidate settles into their minds. Without professional fact-checkers, the burden shifts entirely to the public—and the public, as experts note, is tired, distracted, and already struggling to tell the difference between a genuine photograph and a synthetic one. The result, they predict, will be a sharper degradation of public debate across Latin America, a region where social media is not just a pastime but a primary source of news for millions. It is also a region deeply familiar with disinformation campaigns during major elections, from Brazil to Mexico to Colombia.
Yet there is one notable exception in Meta’s trial: Brazil. The largest market for Meta’s platforms in the region, and one of the most consequential digital arenas in the world, Brazil was left out of the Community Notes experiment. Ivan Paganotti, a professor at the University of Sao Paulo, links this exclusion to Brazil’s far tougher stance on digital platforms. Unlike many of its neighbors, Brazil has regulations that have expanded and forced platforms to develop control and verification mechanisms. The contrast is striking. Elsewhere in Latin America, there is a dearth of legislation governing how social media companies handle false content, leaving platforms free to set their own rules. Brazil, meanwhile, has shown a much stronger willingness to regulate Big Tech. In 2024, its Supreme Court blocked X from operating in the country for 40 days until the platform complied with court orders to remove accounts accused of spreading misinformation. It was a dramatic and unprecedented action, and it sent a signal around the world that Brazil would not let platforms operate above the law. Then in June, in another first for Latin America, a Brazilian court tightened regulations on digital platforms regarding their role in spreading disinformation and hate speech. Brazil’s presidential elections are scheduled for October, and the stakes could not be higher. Regulators there have learned from past experience, including the disinformation campaigns that marred previous votes. By excluding Brazil from Meta’s trial, the company may be acknowledging, without saying so aloud, that it knows crowdsourced moderation cannot withstand the pressures of a high-stakes election in one of the world’s most digitally connected nations. It is an implicit admission that Community Notes are not a reliable safety net, at least when the consequences are too dangerous to risk.
The timing of Meta’s shift adds another layer of unease. Just as the company moves away from professional fact-checking, generative artificial intelligence has evolved at a dizzying pace, producing images, audio, and video that are increasingly difficult to distinguish from reality. Marco Benalcazar, director of the Alan Turing Artificial Intelligence and Vision Research Laboratory at Ecuador’s National Polytechnic School, initially saw the idea of a community helping provide context as positive. But he has become troubled by studies showing that around half of users cannot tell at a glance whether content is real. That number, he said, will grow as generative AI tools advance. Someone with a smartphone and free software can now create a fake video of a political leader saying something they never said, in perfect Spanish, with realistic gestures. What happens when that video circulates on Instagram, and the only safety mechanism is a Community Notes feature that relies on volunteers to flag it? The note may come hours later, after the video has already been shared thousands of times. By then, the damage is done. Martin del Campo agreed that generative AI will be a challenge, and he stressed that basic digital literacy skills are now more indispensable than ever. But asking the public to become better consumers of information feels like an unequal fight when corporations are simultaneously dismantling the very structures designed to guide them. Studies have consistently shown that fact-checking can help curb the spread of misinformation and help users identify false narratives. It works in tandem with labels, warnings, and down-ranking. Without that infrastructure, even the most vigilant users are left to navigate a digital wilderness alone. The irony is that this moment of maximum technological complexity—when deepfakes are becoming cheap and easy to produce—is exactly when platforms should be doubling down on verification, not abandoning it.
The most unsettling possibility, according to analysts, is that Meta’s move may be the beginning of a broader trend. Other social media platforms are watching carefully. Fact-checking is expensive, and it has become politically contentious, with some conservatives accusing these programs of censorship. By stepping back, Meta and X are moving away from verification without facing serious opposition from governments, their users, or public opinion, said Rafael Rubio, co-director of the Complutense University of Madrid’s Disinformation Observatory. That reality, he said, will probably encourage other platforms to do the same. In a profit-driven industry, why invest in teams of expert fact-checkers if you can hand the responsibility to unpaid users and call it community empowerment? Why risk criticism from politicians when you can pass the buck to the crowd? The consequences of this race to the bottom could be profound. If the professional verification model collapses, independent fact-checking agencies may be forced to shrink or disappear. They are funded partly by platforms, and as those partnerships end, their ability to do public-interest journalism diminishes. Meanwhile, the very concept of a shared, verifiable truth comes under renewed assault. For Latin America, a region that has fought hard to build democratic institutions and accountable media, this is a moment of real peril. The conversation is not really about staffing or moderation algorithms. It is about whether the networks that carry our public conversations will continue to value accuracy, or whether they will become indifferent spaces where anything can be said as long as it performs well. The experts who spoke to AFP are not opposed to community involvement in moderation. They simply believe it should supplement, not replace, the people who do this work professionally. As the mud settles, the question left hanging in the air is simple: in a world full of lies, who will still be paid to tell the truth?