IN THIS EDITION
ON THE RISE
Weekly Watch of Emerging Disinformation Risks
So much the worse for the facts?
One of the most remarkable features of disinformation is its ability to maintain the appearance of being right even when reality proves otherwise. Russian disinformation, for instance, routinely fabricates (false) evidence of NATO troops being present in Ukraine because the facts did not fit the narrative it wanted to advance (“Moscow is fighting against the whole Alliance”). US disinformation, meanwhile, continues to repeat the claim that the European Union censors free speech, despite the State Department office tasked with investigating whether the EU’s Digital Services Act (DSA) was being used for censorship finding no evidence to support that claim after weeks of research. The pattern is always the same. “If the facts do not conform to the theory, so much the worse for the facts.” Disinformation will create “alternative facts”, in the form of statements, news stories, images and videos.
Of course, domestic actors spreading disinformation use the same technique. Whether the topic is vaccines, climate change or racism, when reality itself disproves conspiracy theories, the internet and social media provide the perfect habitat for a parallel reality in which those same theories are instead confirmed, with “evidence” to prove it.
The ongoing FIFA World Cup in the United States, and in particular the match between France and Morocco – considered a “high risk” event – was simply too tempting an opportunity for those who spread racist and Islamophobic disinformation (whether driven by ideology or profit) and they did not let it go to waste. In several countries and in multiple languages, including non-European ones, videos showing riots and violence in Paris circulated alongside misleading captions falsely linking them to France’s victory in the match. The only problem: no such unrest had taken place, and the situation in the capital was largely calm. The videos being circulated were old footage, taken from unrelated events.
An unregulated information environment like social media – which, according to the latest Digital News Report by the Reuters Institute, has become the main source of news for many people – is the perfect place to distort reality (even more so now that AI is capable of generating increasingly realistic images and videos). The lack of editorial responsibility among platforms effectively shifts the burden of responsibility onto users. And if, as has already been argued, those who create disinformation face little to no effective deterrence (quite the opposite, in many cases), those who consume it are often, through no fault of their own, poorly equipped to defend themselves.
Confirmation bias (the cognitive bias that leads us to interpret information in ways that align with our existing opinions and belief patterns) and cognitive laziness (our tendency to rely on mental shortcuts, such as accepting familiar information, seeking confirmation of our existing beliefs, or trusting emotionally compelling narratives instead of engaging in deeper analysis) are evolutionary features of our brains that make us particularly vulnerable to disinformation, and can even turn us into its unconscious amplifiers. And if we cannot change the way our brains have evolved over time, then it becomes all the more necessary to demand that safeguards designed to protect our humanity, with all its imperfections, are created and properly enforced.
ZOOM-IN
A Closer Look at Cases Detected by the EDMO Network
The 2026 FIFA World Cup has generated a wide range of disinformation narratives extending well beyond the pitch. The Weekly Pulse of 25 June already warned about the rise of AI-generated content exploiting the tournament’s global visibility, including troves of images of hyper-sexualized female supporters of national teams, designed to amass views, likes and shares, and profit from the social media platforms’ monetization programs. Further analysis by the EDMO Network confirms the pervasiveness of World Cup-related deepfakes, alongside older manipulation techniques such as decontextualized images or videos. Content falsely attributing statements to FIFA officials, false stories about key players and baseless allegations of match-fixing and corruption have been detected, in addition to fabricated stories aimed at spreading racist and Islamophobic narratives.
Racism is the main form of abuse on social media during the World Cup according to FIFA
Chequeado examines the rise of racist abuse during the 2026 FIFA World Cup, highlighting that racism was the most common form of discrimination detected by FIFA’s Social Media Protection Service (SMPS). According to FIFA, more than six million social media posts and comments were monitored during the group stage, with 89,000 identified as abusive. Of these, 11% contained racist content, an increase from 8% during the equivalent stage of the 2022 World Cup. FIFA’s moderation system, which combines artificial intelligence with human review, automatically hid more than 181,000 offensive comments and referred around 1,000 accounts for further investigation. The FIFA Social Media Protection Service is constantly developing. A novelty in this edition is that it has collected data from the security forces, which has made it possible to identify 100 legally attributable cases throughout the group stage of the World Cup.
Chequeado also discusses several high-profile racist remarks made by public officials, including Argentine presidential adviser Santiago Caputo, Mendoza vice-governor Hebe Casado, and Paraguayan senator Celeste Amarilla, who targeted players such as Kylian Mbappé. Experts interviewed by Chequeado argue that these incidents reflect broader societal patterns rather than isolated events. They warn that racist language in football becomes especially harmful when it is normalized or amplified by political leaders, reinforcing discriminatory attitudes and legitimizing prejudice both online and offline.
Algeria-Austria match supposedly “fixed”: why these unfounded accusations?
The article examines unfounded allegations that the 3–3 draw between Algeria and Austria in the final round of the 2026 FIFA World Cup group stage was “fixed.” Because the result allowed both teams to qualify for the knockout phase, numerous social media users claimed, without evidence, that the two federations had colluded. Posts circulated on X linking a video of an on-field confrontation and presenting it misleadingly as proof of a prearranged outcome. Other posts shared on Facebook and Instagram presented a fabricated FIFA press release claiming the governing body had opened an investigation.
The fact-check also notes that many of the posts spreading the rumour originated from Moroccan supporter accounts, suggesting a coordinated disinformation campaign rather than organic commentary. To provide context, investigative journalist Romain Molina explains that although match-fixing still exists, it is highly unlikely at the World Cup because of the tournament’s intense scrutiny and high stakes. The article recalls historical examples of proven collusion, including the infamous 1982 West Germany–Austria “Disgrace of Gijón” and several manipulated 2010 World Cup qualifiers and warm-up matches, emphasizing that past incidents do not substantiate accusations against the 2026 Algeria–Austria match.
Infantino did not say that the president of Mexico offered him money to win the World Cup: the video was created with AI.
A viral video claiming FIFA president Gianni Infantino said Mexico’s president, Claudia Sheinbaum, offered him money to ensure Mexico won the 2026 FIFA World Cup is an AI-generated deepfake, according to Newtral’s fact-check. The video fabricates both the audio and the apparent lip movements. In the clip, Infantino supposedly says he rejected the alleged bribe and that “the only World Cup winner is Messi.”
Newtral found no evidence that Infantino ever made these remarks. Instead, forensic analysis using AI-detection tools, including Hiya (via InVid) and Hive AI, indicated a high probability that the audio was artificially generated, while the visuals were digitally altered to match the fake speech. Reverse image searches traced the original footage to a genuine 2022 video in which Infantino discussed violence at an Argentine football match, not Mexico, Sheinbaum, or the World Cup. The article also notes that some TikTok users labeled the content as AI-generated.
GLOBAL PULSE
Disinformation narratives shaping the world’s conversations
AI funeral images inflate support for Iran’s former Supreme Leader
As Iran commemorated former Supreme Leader Ali Khamenei, several widely shared images and videos purported to show enormous crowds gathering for the funeral ceremonies. However, fact‑checkers found that many of the most viral visuals had been generated or manipulated with artificial intelligence to exaggerate attendance.
Investigators at 20 minutes, a member of DE FACTO, identified numerous warning signs, including unrealistic crowd patterns, perfectly aligned banners, altered landmarks and inconsistencies in the layout of buildings. Some videos also showed impossible proportions, static crowds, or other visual anomalies typical of AI‑generated content.
US President shares old, AI-manipulated photo as evidence of new US strike on Iran, warning it can “get much worse”
A dramatic image shared online was neither recent nor authentic. Fact-checkers traced it to a 2025 photograph of a fire at an oil depot in Tehran, unrelated to the events it was being used to depict. The photo was circulated by prominent accounts and even reposted by US President Donald Trump.
The image had also been digitally altered, with visible elements modified compared to the original photograph. Although the account that first shared it later acknowledged the mistake, the image had already spread widely and was used to support claims about ongoing military operations. The incident shows how easily old images can be repurposed during rapidly evolving conflicts, especially when enhanced with AI tools that make them appear more dramatic and convincing.
Anti NATO disinformation surged around Ankara summit
As NATO leaders gathered in Ankara in July 2026, researchers observed a sharp increase in anti‑NATO narratives targeting Romanian audiences. An analysis of more than 1,000 articles and thousands of Facebook posts found that content linked to the pro‑Kremlin Pravda Romania network intensified significantly during the summit period.
Rather than simply reporting on the event, these outlets promoted a consistent set of narratives: portraying NATO as responsible for escalating tensions, depicting the Alliance as internally divided, framing defence measures as provocations, and suggesting that recent drone incidents in Romania were deliberate Ukrainian attacks.
Researchers also found that such narratives did not remain confined to fringe channels. Similar or identical claims appeared on Romanian media websites, allowing messages originating from low‑reach disinformation networks to gain broader visibility during a key moment in European security debates.
Wildfire conspiracy theories target renewable energy projects
Following the devastating wildfire in Los Gallardos, Spain, social media posts claimed the blaze was deliberately set to clear the way for a large solar energy project. Fact‑checkers found no evidence for this allegation: the projects cited in viral posts were either located outside the affected area or were no longer viable.
The claims mirror a recurring narrative seen elsewhere in Europe. A recent Greek fact‑check similarly debunked allegations that wind farms were being built on land destroyed by wildfires. In both cases, genuine renewable‑energy projects were falsely linked to major fires, turning public concerns about energy infrastructure into unsupported conspiracy theories.
ON A DIFFERENT NOTE
AI has not completely altered the fundamental logic of disinformation, but has significantly increased its scale, speed and accessibility. Deepfakes and AI generated content are not consistently more convincing than traditional forms of disinformation, but generative tools make it easier and cheaper for a wide range of actors to produce and disseminate misleading content. (…)
Only a long term, collective effort by platforms, media professionals, factcheckers, researchers and policymakers – with continuous monitoring, systematic experimentation and adaptation of practices – can keep democratic information provision sufficiently resilient.
Paolo Cesarini, Editorial Director
Tommaso Canetta, Editor-in Chief
Editorial Staff include Elena Coden, Paula Gori, Elena Maggi
This edition draws in part on automated translation and reflects information available as of 15 July 2026. Later developments may not be included.
