⚡ The 30-second brief

• The debunk: Forensic analysis confirms viral images of Zohran Mamdani and Mira Nair with Jeffrey Epstein are AI-generated deepfakes originally posted by a parody account.

• New tech: A new AI-powered tool designed to detect and alert users to malicious, manipulated content across major social platforms.

• The intelligence stack: Using Overpass Turbo to filter and find specific landmarks or infrastructure around the world.

🔦 In the Spotlight

• Fact-check: The release of a new tranche of files related to convicted paedophile Jeffrey Epstein sparked an avalanche of social media discussion that brought with it plenty of misinformation. Fakes included images of a young Zohran Mamdani with his mother in the company of the financier, as well as doctored email screenshots purportedly between Elon Musk and Ghislaine Maxwell. BBC Verify tracked the story in its live coverage.

• Monitoring: We covered the rising tensions between the U.S. and Iran last week, but the situation ramped up further on Tuesday when the U.S. shot down an Iranian drone that it said was approaching the USS Abraham Lincoln in the Arabian Sea. See our OSINT guide for tips on how to follow this story.

🖋️ The Lead: AI-generated images fabricate Epstein connection

In the wake of the latest release of Jeffrey Epstein documents, viral images surfaced appearing to show New York City Mayor Zohran Mamdani and his mother, director Mira Nair, alongside Epstein. However, forensic analysis by VRT NWS, BBC Verify, and France24 confirmed these images are AI-generated deepfakes.

Key Findings

• Source of disinformation: The images feature a “DFF” watermark, traced to the X account @DumbFckFinder, a self-described parody profile that shares AI memes. Despite their satirical origin, the images are being weaponized by political opponents. (VRT)

• Technical verification: Using Google’s tool confirmed that the images were generated using its AI models. No authentic photographs of Mamdani or Nair with Epstein exist in the official case files. (Shayan Sardarizadeh)

• The grain of truth: Mira Nair’s name appears only once in the millions of released pages—not as a co-conspirator, but in a 2009 email regarding a film afterparty for Amelia. There is no evidence of a personal relationship or meeting. (France 24)

Epstein file resources

📍Epstein Library | United States Department of Justice: The website is a repository of millions of documents, images, and videos collected by the FBI and DOJ over decades of investigations.

📍FBI Records: The Vault — Jeffrey Epstein: A repository of declassified and released documents concerning Jeffrey Epstein.

📍Jmail: A website simulating the user interface of Jeffrey Epstein’s inbox and Drive in a searchable format.

— Maria and Derek

⚙️ The intelligence stack

🔍 Overpass turbo
While many journalists use basic satellite imagery, Overpass Turbo allows for a "query-based" approach to geolocation that is essential for verifying footage from conflict zones or remote areas where landmarks are sparse.

Instead of manually scanning Google Earth, you can use a simple script to find the exact intersection of specific features (e.g., a “mosque” within 100 meters of a “pharmacy” and a “school”).

How to use it:

• Go to overpass-turbo.eu.

• Type a query like:

/*
This query looks for surveillance cameras 
within 50m of a supermarket.
*/
[out:json][timeout:25];

// 1. Find all supermarkets in the current map view
node["shop"="supermarket"]({{bbox}})->.supermarkets;

// 2. Find surveillance cameras near those supermarkets
node(around.supermarkets:50)["man_made"="surveillance"]->.cameras;

// 3. Output the results
(.cameras;);
out body;
>;
out skel qt;

This script searches for all surveillance cameras (man_made=surveillance) located within 50 meters of a supermarket (shop=supermarket). This is highly effective for locating potential CCTV sources for a specific incident.

• Zoom into your target city and hit Run.

💡 Quick tip: If you find the manual Overpass QL code too cumbersome to write, the Wizard is your best friend. The Wizard translates “human-readable” terms into functional queries so you don’t have to worry about brackets or semicolons.

📍GEOINT: Learn how to turn satellite imagery into an empirical audit trail (Eurovision News Spotlight)

📍Satellite metadata: An OpenC3 COSMOS plugin for tracking satellites using the KeepTrack Space API. This plugin provides real-time satellite position tracking and orbital element data. (Clay Genki Kramp)

📍Temporal Analysis Engine: Extract and visualize temporal metadata from any URL. Discover creation dates, modification history, and hidden timestamps. (Roelof T.)

📍Company OSINT: The Coalition of Cyber Investigators discuss some of the ways that OSINT can help manage the risk of blind spots when you have operations in less mature markets. (Coalition of Cyber Investigators)

📍Deepfake Guard: New AI-powered tool designed to detect and alert users to malicious, manipulated content across major social platforms (Avast)

📍OSINT at home: 6 free satellite imagery tools every investigator should know ()

📍U.S. Armada: An OSINT intelligence guide to the naval build-up near Iran (Eurovision News Spotlight)

📍Reconya: Real-time network device discovery and monitoring ()

📡 Signal analysis

In the 15th edition of the Spotlight Review, EBU members joined Maria Flannery to discuss the long-running problem of scam advertisements running on Facebook and the difficulties in verifying content from Iran’s recent protest movement.

Tim Verheyden (VRT) discussed a recent documentary that aired on the Belgian broadcaster’s investigative Pano programme, detailing a Europe-wide problem of industrialised scamming that has seen victims lose out on hundreds of thousands of euro.

Eva Wackenreuther (ORF) detailed how their investigation began in Austria after one of their newscasters was the subject of a deepfake ad on Facebook.

Before we go…

AI unlocks hundreds of cosmic anomalies in Hubble archive

Astronomers using an AI-assisted tool called AnomalyMatch discovered a "hamburger-shaped" planet-forming disk in the Hubble Archive. It's a classic example of automated anomaly detection surfacing something humans missed for 35 years because we weren't looking for "fast food" in deep space.

⚡ The 30-second brief

• The debunk: Sync-verification of eyewitness footage in Minneapolis refuted “domestic terror” claims, proving Alex Pretti was holding a phone, not a weapon, when shot.

• New tech: Emerging research into “collaborative malicious AI agents” reveals how automated swarms now coordinate to fabricate social consensus in real-time.

• OSINT stack: Bypassing Telegram’s search friction using Google Dorks (site:t.me) to extract cached messages and hidden forensic leads.

🔦 In the Spotlight

• Fact-check: Spotlight members were swiftly working to verify different angles of the shooting of Alex Pretti on a Minneapolis street in broad daylight on Saturday. Read a full breakdown by Spain’s RTVE and watch a TV segment by BBC Verify. And while much of the circulating material was real, there were also AI-retouched images being shared. Read more by SRF and 3CatInfo, or listen to this ORF report (German).

• Monitoring: The internet blackout in Iran was still hampering Iranians’ efforts to get videos online, and many of what did emerge showed the scenes in early January when the government began a brutal crackdown. Spotlight members have been looking for new material to help us understand what is happening inside the country, where rights group HRANA is reporting that it has confirmed over 6,000 deaths. They are still investigating a potential 17,000 more deaths.

🖋️ The Lead: Video verification at heart of U.S. ‘inflection point’

The shooting dead of an American citizen by federal agents in Minneapolis at the weekend further stoked anger in the city and sparked outrage across the U.S. Alex Pretti, a 37-year-old nurse, was killed during a confrontation between immigration agents and civilians.

Secretary of Homeland Security, Kristi Noem, accused Pretti of “domestic terrorism”. Other discourse, including from Noem, centred around claims that Pretti brandished a weapon and was about to attack officers. It quickly emerged that Pretti was a legal gun-owner in a state where carrying such weapons is allowed.

Crucially, the video evidence did not back up the ‘domestic terror’ claims. Spotlight members verified multiple clips from the scene, showing Alex Pretti filming with his phone in his hand. After being tackled to the ground, an officer appears to pull Pretti’s gun from his waistband, removing the civilian’s gun from the altercation. Then, Pretti is shot in the back; ten shots ring out in a five-second video.

The result was what Governor Tim Walz called an “inflection point” for America. Now, the U.S. administration has indicated that it may change course in Minnesota, less than three weeks after the killing of Renee Good, a case in which swift video verification also played a key role.

— Maria

⚙️ The intelligence stack

🔍 Advanced search operators for Telegram
Telegram’s native search is notoriously limited, often hiding critical eyewitness media or niche channels from direct view.

💡 Quick tip: When investigating breaking news or monitoring extremist groups, don’t rely on the app’s internal search. Instead, use a targeted dork on a search engine to surface messages, files, and channel invites that haven’t been indexed elsewhere.

The Syntax: site:t.me "keyword" - "view in telegram"

• Why it works: By excluding the phrase “view in telegram,” you filter out the standard channel landing pages and force the search engine to index the actual message content and preview text.

• The OSINT Edge: This approach often reveals “private” channel links or deleted message previews that are still cached in search engine results.

📍Forensic smartphone verification: Master the "zoom-in" lens to reconstruct high-stakes use-of-force events (Nieman Lab)

📍Automated fact-checking: Implementing Google APIs and ClaimReview schema to accelerate OSINT workflows (via Eurovision News Spotlight)

📍Countering AI bot swarms: Forensic strategies for detecting "collaborative malicious AI agents" designed to mimic human consensus (The Guardian)

📍Cross-platform identity stitching: Linking accounts via photo reuse and username structures (ShadowDragon)

📍Climate misinformation forensics: Technical requirements of the new EU Declaration on deceptive climate change narratives (European Commission)

📡 Signal analysis

OSINT Industries has published a new case study detailing how investigators used OSINT and social media footprinting to intercept a tiger cub trafficked for illegal sale.

This investigation provides a rare, technical look at how digital forensics can be applied to wildlife crime, using real-time intelligence to trace supply chains and unmask the coordinators of high-value animal trafficking networks. (via OSINT Industries)

Before we go…

Schweizer Radio und Fernsehen (SRF) is hosting a dedicated “Fact or Fake?” theme week from January 26 to 30, 2026, with reports, investigations, and fact checks on current issues in politics, business, and society.

The broadcaster is inviting the public to submit suspicious media via WhatsApp or web forms for expert verification and investigative reporting.

⚡ The 30-second brief

• The debunk: Debunking the hoaxes surrounding the Adamuz train disaster

• New tech: WebRecon for comprehensive crawling, data extraction, and analysis

• OSINT stack: Reconnaissance tools for data extraction and vibecoding

🔦 In the Spotlight

• Fact-check: AI-generated images of the wreckage were circulating after a deadly train crash in southern Spain. Read the full breakdown from Spain’s VerificaRTVE and 3CatInfo.

• Monitoring: Donald Trump’s continued threats to seize control of Greenland sent European leaders into a tailspin. Denmark and other allied countries sent troops to Greenland to participate in military exercises. With the situation escalating politically, we are monitoring for opportunities to add intelligence through open-source investigations.

🖋️ The Lead: Debunking the hoaxes surrounding the Adamuz train disaster

A crash between two high-speed trains on January 18 in Adamuz, Córdoba, left at least 41 people dead and injured more than 100 people. Once emergency services began to process the scene, official images began to emerge showing the extreme wreckage that was left along the track in the wake of the collision.

Getting a clear assessment of the situation on the ground was complicated by the spread of AI-generated images among the genuine footage. And it wasn’t just images: false claims were also circulating before investigators had even had time to make a full assessment of the circumstances surrounding the crash.

Debunking the fakes:

• Two widely circulated images were found to be AI-generated by comparison against existing official footage from the scene, showing that the trains were not positioned exactly as they were in reality. (One of these AI pictures was even used by a newspaper in its coverage of the crash.)

• A fake message circulated asking the public to help at the scene of the crash, and to come wearing high-visibility vests. The Regional Government of Andalusia confirmed that this was a hoax, and asked civilians not to come to the site.

• Posts claimed that Spain gave €247 million to improve Morocco’s railway network, but that was misleading. In reality, Spain did give Morocco (and Uzbekistan) funding, but in the form of repayable loans. In Morocco’s case, the loan was subject to a Spanish company being awarded a contract for the job.

• Social media posts claimed the trains were speeding, but the national railway company RENFE said the trains were travelling at 205 km/h and 210 km/h, on a section of track with a 250 km/h speed limit.

— Maria

⚙️ The intelligence stack

🔍 WebRecon
WebRecon is an advanced web reconnaissance tool designed for cybersecurity professionals, penetration testers, and security researchers. It automates the process of gathering intelligence from target websites through comprehensive crawling, data extraction, and analysis.

💡 Quick tip: When conducting initial reconnaissance on a high-interest target, avoid triggering security alerts (like WAF blocks) by running a shallow, filtered scan through a proxy.

The Command:

python3 WebRecon.py [TARGET_URL] --max-depth 1 --no-images --proxy socks5://127.0.0.1:9050

📍 Data discovery via vibecoding: Using AI to find hidden data sources (via )

📍Mastering the SIFT method: A practical manual on using lateral reading and search operators to verify digital leads and protect the chain of custody. (via Eurovision News Spotlight)

📍Suspect and car investigations: A technical walkthrough detailing the methodology for linking isolated data points—specifically suspect identities and vehicle movements. (via UK OSINT Community)

📍Kling AI: Generating a synthetic person using Kling’s new motion control engine (via Emmanuelle Saliba)

📍Domain search: Domain data for research purposes without any fee or paywall (via ip.thc.org)

📡 Signal analysis

Why Grok restrictions won’t stop society’s latest AI scourge
Kate McDonald explores how X’s new geoblocking and image-editing restrictions for Grok fail to address the root issue, as “Spicy Mode” and standalone app access continue to facilitate the generation of non-consensual deepfakes. (via RTE Clarity)

Before we go…

UNESCO has published the second edition of Story-Based Inquiry: A Manual for Investigative Journalists, offering a comprehensive update of its 2009 global reference handbook.

This fully updated second edition brings new research, innovative practices and insights from global investigations. (via UNESCO)

⚡ The 30-second brief

• The Debunk: Verified graphic Kahrizak morgue footage as contemporary by geolocation and analysis of visible date indicators, disproving claims it was recycled.

• New Tech: Snowflake (Tor) now masks censorship circumvention as standard WebRTC calls to evade Deep Packet Inspection (DPI).

• OSINT Stack: Sherlock (username tracking), Robin (AI Dark Web analysis), and RageCheck (linguistic manipulation detection).

• Truman Show: An AI‑powered investment scam that uses apps from the official mobile app stores to steal money and identity data from victims.

🔦 In the Spotlight

• Fact-check: Spotlight members collaborated to verify videos showing the fatal shooting by an ICE agent of Renee Nicole Good in Minneapolis. A frame-by-frame analysis challenged the U.S. administration’s claims that she attempted to run an officer over. Read ZDF’s full breakdown.

• Monitoring

  • Iran: A near-total internet blackout continues as the regime attempts to mask a lethal crackdown on nationwide protests triggered by a December currency collapse. Connectivity remains at approximately 1%, with reports verified by the Spotlight network and human rights groups suggesting thousands of casualties and mass arrests since January 8.

  • Greenland: Tensions between Washington and Copenhagen have reached a post-WWII high following President Trump’s Donroe Doctrine declaration. Framing the island as a national security imperative to block Russian and Chinese Arctic expansion, the White House has actively drafted acquisition plans and refused to rule out “the more difficult way” — military action — to secure the territory.

🖋️ The Lead: Piercing the Kahrizak blackout

On January 10, 2026, as a state-mandated internet blackout blanketed Iran, graphic footage began emerging from the Kahrizak district. The videos, appearing to show a makeshift morgue overflowing with victims of a recent crackdown, shocked the diaspora. In a region where large scale protests and crackdowns have taken place repeatedly over recent years, verification was essential.

While the Iranian government maintained a narrative that the unrest was under control, members of the Spotlight network verified the footage as authentic and contemporary.

How we verified the Kahrizak footage:

• The time: We corroborated footage from different sources and saw a date on a screen in one of the clips, proving the scenes were new and not from past protests.

• The place: We confirmed that the buildings in the video are part of the Kahrizak facility using satellite imagery and an architectural rendering of the centre. Logos seen on body bags were linked to Tehran’s main cemetery.

• The disinformation: We tracked state media for references to the situation in Kahrizak; reporters blamed ‘rioters’ for the killings. But Spotlight heard verified voice messages from eyewitnesses confirming a brutal crackdown on protesters.

The takeaway: Even when the internet is cut, the truth leaves a footprint. Although content coming out of Iran is scarce, verification on the little we can access is possible.

🇮🇷 Iran digital forensics field guide

Amid the blackout, the Spotlight network released a definitive digital forensics field guide, transforming how newsrooms can extract evidence from behind the firewall.

The three-step workflow:

• Smuggled data: How to find and verify footage sent via mesh networks that don’t need the internet.

• Network tracing: Using “leaks” in the system to prove exactly where and when the government cut the signal.

• Physical evidence: Using shadows, sun position, and landmarks to verify a video’s location when digital data is missing.

⚙️ The intelligence stack

🔍 Snowflake: A pluggable transport from The Tor Project designed to bypass internet censorship by masking Tor traffic as a WebRTC video or voice call.

It works through a special system where "brokers" connect users who can't access the internet with volunteers using regular web browsers, making it very hard for governments to block it without also affecting real video calls.

💡 Quick tip: For journalists in high-risk zones, if Tor fails, go to Settings > Connection > Bridges > Built-in Bridge and select Snowflake. It is harder to block than standard VPNs because it mimics legitimate teleconferencing.

📍 YouTube investigations: published a detailed look at more than a dozen free tools you can use to investigate YouTube videos, channels, and comments (via Indicator.media)

📍 Urban geometry forensics: Richard Irvine-Brown’s guide to geolocating via road curvature, "barcode" pattern matching, and multi-platform satellite verification. (via @richib.bsky.social‬)

📍 Sherlock: A powerful command-line OSINT tool used to track username footprints across hundreds of websites to uncover a target's digital presence. (via GitHub)

📍 Sentiment analysis for investigators: ’s guide on using NLP to measure changes in emotions and spot organised influence efforts. (via Eurovision News Spotlight)

📍 RageCheck: Open-source pattern detector that scores content for loaded language and us-vs-them divisiveness. (via GitHub)

📍 Robin: AI-powered Dark Web investigator with automated workflows for macOS/Linux. (via GitHub)

📡 Signal analysis

🎧 Minnesota ICE shooting: On The Media breaks down the frame-by-frame analysis of the Renee Nicole Good shooting, challenging official claims of an attempted vehicular assault. (via WNYC Studios)

🏁 Before we go…

The “Truman Show” scam: Trapped in a synthetic reality
Check Point Research has exposed OPCOPRO, a scam that builds a personalized, AI-generated “reality” where every expert and peer in your chat group is a synthetic persona.

The operation is a fully synthetic, AI‑powered investment scam that uses legitimate Android and iOS apps from the official app stores and AI‑generated communities to steal victims' money and identity data. (via Check Point)

⚡ The 30-second brief

• The debunk: Viral images of Nicolás Maduro’s capture were AI fakes. We identify the three specific ‘artefacts’ that fooled millions.

• New tech: Google’s SynthID has moved from beta to Gemini; we show you how to use it for instant verification.

• OSINT stack: Professional guides on scraping evidence, finding deleted posts from politicians, and tracking vehicle licence plates.

• Data culture: How Radiohead used Spotify metrics to ditch their hits for deep cuts on the 2025 tour.

🔦 In the Spotlight

• Fact-check: We recently dissected a viral image claiming to show Nicolás Maduro being taken into custody by U.S. forces. Read the full breakdown.

• Monitoring: Developments in Venezuela led to heightened speculation about the Trump administration’s potential plans to take military action in other countries, like Greenland, Colombia, Mexico and Cuba. As world leaders prepared their responses, the discourse on social media laid the groundwork for additional OSINT surveillance.

🖋️ The Lead: Seeing isn’t believing in Caracas

On January 3, 2026, a high-resolution image of Nicolás Maduro being escorted to a transport plane dominated global feeds. It was the perfect smoking gun — except it was synthetic.

While the world waited for official confirmation, members of the Spotlight network used Google DeepMind’s SynthID to debunk the imagery hours before the White House released an actual photo of Maduro in custody.

Our digital analysis identified several ‘AI artefacts’ that exposed the deception:

• Impossible physics: The aircraft featured impossible aerodynamics and non-standard landing gear.

• Uniform gibberish: Uniform patches on the guards were nonsensical and matched no known military units.

• Shadow logic: Distorted hand geometry and inconsistent facial shadows confirmed synthetic generation.

The takeaway: The first photo of a major breaking news event is now more likely to be fake than real. Verification must happen at the point of consumption, not hours later.

— Maria

⚙️ The intelligence stack

🔍 SynthID: Google DeepMind’s SynthID Detector was launched in mid-2025 and works by detecting an imperceptible embedded watermark in content generated or altered by Google’s AI programmes. While a SynthID dashboard giving detailed information is open to early testers, the detector has also been integrated into Google’s Gemini AI assistant.

💡 Quick tip: SynthID can work even when content has been somewhat transformed, but heavy filters or photos of the image on a screen can limit its usefulness. In case of an inconclusive result, try to find other cleaner versions of the same image to test.

📍 Mastering web scraping: ’s guide on automating evidence collection and SHA-256 data preservation. (via Eurovision News Spotlight)

📍 Investigative toolkit: Rowan Philp (GIJN) curates 2025’s best tools for tracking corporate misconduct and automating YouTube summaries. (via )

📍Ad fraud & OSINT: discusses tracking the money behind disinformation. (via The 404 Media Podcast)

📍 Migration tracker: BBC Verify’s new postcode-searchable database to contextualize UK migration stats with embeddable modules. (via BBC Verify)

📍 Ghost posts: ’s masterclass on using Factba.se and Politwoops to recover deleted political tweets. (via Bendobrown)

📍 License plate tracking: A new primer on vehicle OSINT using Platesmania and regional databases. (via AirCorridor)

📡 Signal analysis

🎧 FACTS IN : FACTS OUT: Is AI breaking the news? visits the BBC's headquarters in London to dig into cutting-edge research on the growing use of AI assistants such as ChatGPT and Perplexity for news-related questions. (via Eurovision News Podcast)

🎸 Before we go…

Radiohead by the numbers: Investigative skills aren’t just for conflict zones. Richard Irvine-Brown applied an open-source data lens to Radiohead’s 2025 European tour.

By scraping and cross-referencing Spotify streams against RateYourMusic scores, the project reveals a fascinating trend: the band’s 2025 setlists systematically prioritized ‘fan-favourite’ deep cuts over mainstream hits like “Creep”. It’s a masterclass in using cultural data to track a shift in artistic strategy. (via @richib.bsky.social)

Check out our top picks from the year that was.

Tracking Israel’s state-sponsored information campaigns

The Spotlight network conducted extensive research this year on the digital infrastructure used by the Israeli government to distribute advertisements in international territories.

In September, an OSINT reporting group published findings on coordinated messaging campaigns aimed at shaping global perspectives on Gaza, noting contracts valued at tens of millions of dollars. Among the messaging identified were claims that famine conditions did not exist in Gaza — a point of contention following an official Integrated Food Security Phase Classification (IPC) report that had declared famine conditions weeks prior.

While the Israeli government criticized the IPC methodology as “tendentious,” the Spotlight network documented the use of paid advertising to prioritize the state’s rebuttal over international humanitarian data.

Earlier in the year, the network investigated advertisements run by the state agency LAPAM to encourage public voting for the Israeli Eurovision Song Contest entry. This documentation was later referenced in reporting when contest organizers announced reforms designed to maintain the competition’s integrity.

The most recent report on these communication systems identified a new digital campaign following a fragile ceasefire. These advertisements appeared on social media accounts that did not initially disclose government affiliation; verification teams found that some video segments used in the campaign were archive footage unrelated to the current ceasefire.

Analysis of anti-migrant sentiment in Europe

An analysis of the surge in anti-migrant incidents across Europe identified recurring patterns in the digital information ecosystem, where cross-border social media networks frequently amplified local tensions.

In 2025, violent clashes occurred in Spain, the UK, and Ireland. Reporting indicated a broadening of narratives that have historically existed in other European nations, such as Germany. Documentation showed that local influencers — including Tommy Robinson and accounts like ‘RadioGenoa’ — coordinated communication across territories. In Hungary, researchers noted that similar rhetoric was prominently featured in official government communications.

COP30 and narratives on climate action

An analysis of 70,000 social media posts regarding climate action identified a shift in rhetoric within climate-skeptic circles. Researchers observed a move away from challenging established climate science toward emphasizing the economic costs of green transitions and the financial impact on consumers.

During the conference in November, digital discourse focused on the logistics of the event, including the use of cruise ships for accommodation and the opening of a new motorway in Belém. These topics were frequently used to frame the event as “elitist” or inconsistent with environmental goals.

Israel and Iran’s ‘Twelve-Day War’

In June, a direct military confrontation began following an Israeli strike on military and nuclear infrastructure in Iran. Iran responded by launching several hundred missiles and drones toward Israeli population centres.

The United States subsequently conducted Operation Midnight Hammer, deploying B-2 Spirit stealth bombers to strike three Iranian nuclear facilities at Fordow, Natanz, and Isfahan. Iran retaliated with further missile strikes, including one targeting the Al Udeid Air Base in Qatar. A ceasefire was reached shortly thereafter on June 24.

The rapid escalation generated a high volume of online engagement. Fact-checkers identified that much of the visual material circulating during the crisis (intended to depict the conflict) was either AI-generated or recycled footage from more than 20 years ago.

Pope’s death reignites Vatican conspiracy theories

Following the illness of Pope Francis in February, a large volume of AI-generated images appeared online depicting the pontiff in a hospital setting. This trend expanded to include composite images of deceased public figures alongside religious icons.

When Pope Francis died in April, the event led to a resurgence of unverified claims and conspiracy theories regarding the Catholic Church. These included claims of hidden messages and symbols in Vatican architecture and art.

The death of a pope kicks off a long process that holds a certain mystique for election-watchers. The 2025 conclave was perhaps even more popular as it came soon after a hit film of the same name, which delved into the process, the politics and the intrigue of the election process.

To counter the claims of “prophecies” and speculation that had gained traction due to popular culture portrayals of the papacy, verification teams detailed the formal procedures of the 2025 conclave and clarified the sequestered voting process for cardinals.

World Cup tickets and extreme markups

Despite initial commitments to an accessible 2026 World Cup, an analysis of the official resale platform by Eurovision News Spotlight revealed a lack of price capping. Standard stadium seats were observed at prices exceeding $100,000, with one listing reaching nearly $1 million.

FIFA officials stated that their pricing model reflects “existing market practice” in North America and that fees are aligned with industry trends. However, the network’s investigation into every tier price for every match revealed that FIFA stands to gain from the sizeable commission it charges to both buyers and sellers on the platform.

In response to widespread criticism, FIFA later introduced a “Supporter Entry Tier” with fixed $60 tickets, though fans’ groups argued this allocated inventory was insufficient to address broader affordability concerns.

Eurovision News Spotlight collaborations in 2025 featured the work of: Derek Bowler, Maria Flannery, Sara Badilini, Jenny Hauser, Borja Díaz-Merry, Lili Rutai, Kate McDonald, Carlos Baraibar Padró, Jack McCarron, Karen Bauer, Jana Heigl, Max Gilbert, Kathrin Wesolowski, Michael Schlegel, Eva Wackenreuther, Alima de Graaf, Daan Nicolay, Mahsa Aminolahi, Satu Helin, Ville Juutilainen, Giles Pitts, Johanna Vehkoo, Sarah Elmer, Blanca Bayo Pérez, Uta Steinwehr, Milena Aellig, Grégoire Ryckmans, Estefanía de Antonio, Camille Pettineo, Anita Fichtinger, Justyna Kurczabinska, Gunnar Krüger, Sarah Schmidt, Alice Palussiere, and Valentine Joubin.

The incident — which has been declared a terror attack by local police and is believed to have been inspired by Islamic State — was allegedly carried out by a father and son duo. Sajid Akram, aged 50, died at the scene, while 24-year-old Naveed was in hospital in critical condition.

A 10-year-old girl, a Holocaust survivor, a rabbi and a retired police officer who was working as a photographer at the event were among those killed.

As the attack unfolded in Sydney, social media platforms were flooded with speculation and unverified claims with no basis in fact. This trend of misinformation spreading on social media platforms following major incidents has been frequently observed.

The story of the heroic actions of a bystander who wrestled a gun from one of the shooters was distorted, when a false name circulated along with claims that he was trending on Google before the incident even happened.

As we have seen in previous terror-related incidents, there were also claims that it was a false flag attack.

Here, we have collated some of the most pervasive claims and provided a fact-check analysis of the content.

Hero of Bondi Beach falsely named as ‘Edward Crabtree’

One of the most viral videos from the incident showed a man confronting one of the gunmen, wrestling the weapon from his hands, and turning the gun on the shooter.

The clip was verified by numerous large news organisations and by authorities in Sydney, who called the man a “genuine hero” who may have saved many lives during the mass shooting.

The man who wrestled the gun from the shooter, named by local news outlets as 43-year-old fruit shop owner Ahmed al Ahmed, suffered gunshot wounds in the incident and was receiving treatment in hospital.

However, in the hours after the incident, rumours circulated online that the man’s name was Edward Crabtree. It appears to have started on a fake news website appearing to be a national Australian outlet called ‘The Daily’.

An article written by ‘Rebecca Chen’, described as the Senior Crime Reporter, was titled: “‘I Just Acted’: Bondi Local Edward Crabtree Disarms Gunman in Terrifying Attack”.

The article goes on to describe Crabtree as a 43-year-old IT professional and describes in detail how when he “left his Bondi apartment on Saturday afternoon for his usual weekend walk along the beachfront, he had no idea he would soon be facing down an armed terrorist”.

Described as an exclusive interview from his hospital bed, the article quotes ‘Crabtree’ on his thoughts and feelings when he saw the gunman. The claim about the man’s name circulated widely on social media, and was even picked up by X’s built-in AI tool, Grok, which repeated the misinformation when asked who the man was.

Despite how widely the name circulated, subsequent investigation revealed the origin of the false claim. Marie-Laure Mathot from Belgium’s RTBF Fakey dug into the background on the ‘news’ website, and found that several clues indicated that it was potentially created using AI. Among the only stable articles on the site was the one about the Bondi shooting, and the byline photo for Rebecca Chen was changing with each refresh.

But the most conclusive piece of evidence came via a domain name lookup website, Whois, that showed the URL for the news site was created on the day of the mass shooting. The data of those who created the domain name was hidden behind a privacy service in Reykjavik.

Grok gets it wrong again

The name of the heroic passerby was not the only detail Grok got wrong as people searched for information online.

A fact-check by Katja Belousova for Germany’s ZDFheute detailed how the X AI initially told users wondering about the authenticity of the viral video that it was false.

The chatbot said: “This appears to be an old viral video of a man climbing a palm tree in a parking lot, possibly to trim it, resulting in a branch falling and damaging a parked car. Searches across sources show no verified location, date, or injuries. It may be staged; authenticity is uncertain.”

The video was verified by numerous news organisations and formed the basis for a major news angle as journalists and authorities scrambled to learn the truth about the shooting.

This example shows the limitations of LLMs for verification work and should serve as a reminder that, while they are trained on online content and discussions, they are ill-equipped to perform the same level of verification and quality checks on that information as a human analyst.

Any issues with Grok specifically could also come under more scrutiny since an announcement earlier this year that it would be the AI tool of choice for the U.S. government. The Pentagon put pen to paper on a multimillion dollar deal to roll out Grok for government use, the Department of Defense confirmed in July.

Claims that names spiked on Google before attack

Among the baseless rumours that this was a false flag attack were claims that one of the shooter’s names had spiked as a Google search term before the attack even took place.

More specific rumours centred on how the name was being used in Google searches in Australia and Israel before the attack happened, suggesting involved parties.

However, an analysis of Google Trends showed that in Australia, the name started trending at about 9 a.m. GMT on Sunday. The first reports of an active shooter on the beach were at 7:45 a.m. GMT (6:45 p.m. local time).

In Israel, it was the same story. The search term ‘naveed akram’ began trending at 10 a.m. GMT, an hour later than Australia, reflecting the delay it took for news of the name of one of the suspects to go international.

Similar claims emerge frequently amid major incidents, and usually involve the uploaders making a mistake around time zones.

On Google Trends, data is shown in the viewer’s local time, not the local time of the country where the event took place. For an incident in Australia, which has a very different time zone from most of the West, there is even more room for time-related errors.

The claim in the X posts seen above had hundreds of thousands of views alone, and were also spread by other accounts via viral posts, including on TikTok, where some content had since been removed.

Fact-checkers at ORF’s Zeit im Bild also covered this story while tracking Bondi-related misinformation for Austrian TV.

@zeitimbild

Zeit im Bild - ZIB on Instagram: "Fake News im Umlauf: Rund um …

Meanwhile, AI was also used to create images to support the claim that the attack was staged by ‘crisis actors’. An AI-generated image showed a man having blood applied by a makeup artist, VerificaRTVE’s Mario Pérez reported.

AI detectors indicated a high likelihood that the image was AI-generated. But there were also some visual clues: the text on his T-shirt was distorted — in a similar style to the way generative AI often struggles with text.

Man wrongly identified as alleged shooter

Another video that was viral in the aftermath of the attack showed a man claiming his photo was falsely linked to the Bondi attack, and his claims were indeed true, according to DW Fact Check’s Anwar Ashraf and Silja Thoms.

The man lives in Sydney and has the same first and last name as one of the alleged shooters, but there is no connection between him and the attack.

The photo of the man was being shared around after people searched for the alleged gunman on Facebook and found the unrelated man’s page. DW reported that photos of the suspect and images of this man show they are significantly different.

Furthermore, while the older shooter died at the scene, the younger gunman suffered critical injuries and was in hospital. “It is therefore impossible for the man speaking in the video to be the alleged attacker,” Ashraf and Thoms write for DW.

The proliferation of these unverified claims and the subsequent false identification of an innocent person underscore the critical importance of seeking information from verified news reports in the immediate aftermath of major incidents.

SOURCES

• thedailyaus.world (2025). ‘I Just Acted’: Bondi Hero Edward Crabtree Disarms Terrorist - The Da…. [online] [Accessed 14 Dec. 2025].

• Mathot, Marie-Laure (2025). Ahmed al Ahmed ou Edward Crabtree : un faux site d’info créé de toutes pièces pour désinformer sur l’identité du ‘héros’ de Bondi Beach - RTBF Actus. [online] RTBF. [Accessed 15 Dec. 2025].

• ‌X (formerly Twitter). (2025). Available at: Edward Crabtree is a 43-year-old IT professional and senior solutions architect from Sydney, Australia… [Accessed 15 Dec. 2025].

• ‌Whois.com. (2025). Whois thedailyaus.world. [online] [Accessed 15 Dec. 2025].

• Katja Belousova (2025). Bondi Beach: Diese Falschinformationen kursieren zur Attacke. [online] ZDFheute. [Accessed 15 Dec. 2025].

• X (formerly Twitter). (2025). Available at: This appears to be an old viral video of a man climbing a palm tree in a parking lot, possibly to trim it… [Accessed 15 Dec. 2025].

• Jamali, L. (2025). Musk’s Grok signs $200m deal with Pentagon just days after antisemitism row. [online] Bbc.com. [Accessed 15 Dec. 2025].

• Dervisevic, H. (2025). What we know so far about the Bondi Beach shooting. [online] Abc.net.au. [Accessed 16 Dec. 2025].

• X (formerly Twitter). (2025). Available at: BREAKING: BONDI BEACH SHOOTER NAME SPIKED ON GOOGLE TRENDS HOURS BEFORE ATTACK [Accessed 16 Dec. 2025].

• Instagram. (2017). Zeit im Bild - ZIB on Instagram: “Fake News im Umlauf: Rund um den antisemitischen Anschlag in Sydney kursieren viele Falschmeldungen. Auch rund um den „Helden von Sydney‘ werden bewusst Lügen verbreitet.’ [online] [Accessed 16 Dec. 2025].

• Ashraf, A. and Thoms, S. (2025). Fact check: Fake news spreads after Bondi Beach shooting. [online] dw.com. [Accessed 16 Dec. 2025].

• X (formerly Twitter). (2025). Available at: Exposing propaganda Naveed Akram from Lahore who lives in Sydney came on camera to prove his identity and how some media platforms put his life in danger by circulating his picture as alleged… [Accessed 16 Dec. 2025].

• VerificaRTVE and Pérez, M. (2025). Bulos y desinformación sobre el atentado contra una celebración judía en Sídney. RTVE.es. [online] [Accessed 16 Dec. 2025].

  ‌

  ‌

The move meant the army was ousted from its last stronghold in western Darfur. It also highlighted the humanitarian disaster that has long been unfolding in the country, as the RSF were accused of mass killings and other violence in areas they had taken over.

The dire situation was highlighted due to analysis of satellite images by the Humanitarian Research Lab at the Yale School of Public Health, which appeared to show clusters of objects and dark stains on the earth; dead bodies and blood, visible from Space.

Activists have consistently raised concerns about Sudan. Testimony from eyewitnesses has been patchy amid unreliable connectivity, and journalists and humanitarians have struggled to gain access to the affected territories amid widespread looting and violence.

Martin Griffiths, the UN Under-Secretary-General for Humanitarian Affairs and Emergency Relief Coordinator, said in early 2024 that it had been “very, very difficult to get attention to Sudan, which, in my view, is a place of as great a suffering as anywhere in the world today.” The crisis in the country has only deepened since then.

Another angle of this story is a renewed scrutiny of the United Arab Emirates, which has been accused of supporting the RSF in the civil war. In an announcement in early 2025, the U.S. declared that the RSF had committed genocide during the war. The UAE has refuted allegations of arming or otherwise facilitating the paramilitary group.

Precise figures on the number of people killed in the conflict are difficult to pinpoint, but a former U.S. envoy to Sudan estimated the death toll at over 400,000.

Journalists around the world, hoping to cover this story accurately and sensitively, are relying on the few pictures emerging from the affected areas. The vast gap between the demand for content and what is actually available has created the perfect environment for fake and misrepresented content to spread, purporting to show the situation in Sudan.

Misrepresented and AI-generated content

Some of the most viral content around the conflict in Sudan does not actually show what it claims, writes Matt Ford for Deutsche Welle in an analysis of various videos and images.

One video was widely circulated showing a mother and her three children, with captions saying that she was cowering in front of RSF troops who were allegedly firing weapons near her.

The footage was reshared repeatedly by large accounts. The story of the woman grew more layers when posters claimed it showed the situation as el-Fasher was captured by the RSF, and added that the mother had been killed without any evidence.

By running a reverse image search, Ford found that the video had been posted in mid-September, more than a month before el-Fasher fell to the RSF. Local language experts at DW’s Arabic desk analysed the dialogue in the footage, which suggested the men were not members of the RSF but were actually from the Sudanese Armed Forces. “They don’t appear to be threatening the woman but are rather asking her where she is from and whether she is linked to the RSF. The woman replies that she is part of the Zaghawa tribe and that the RSF arrested her husband,” Ford writes.

Other videos were also circulating. For VRT’s nws check, Ellen Debackere and Daan Nicolay investigate several clips spreading across social media, writing that, unfortunately, many gruesome images were authentic, but AI-generated material was also present.

A photo purportedly depicted a mother and child just before their execution by RSF fighters. However, there were several indications that it was actually AI-generated, Debackere and Nicolay write.

First, a small logo in the image and a reverse image search both led to the Instagram accounts of a “digital video producer” and a “creative AI specialist”. All circulating versions of the image bore his logo.

“The creator himself indicates in some places that the images were created using AI. For example, in the caption of the images on his Instagram channel, he writes in Arabic that the images were created using AI. He also includes an AI icon in an Instagram story of the image.” A video showing the same scene also had some inconsistencies, such as the merging of the woman’s arms.

Another viral image covered by VRT was also unpicked by the team at VerificaRTVE in Spain. The image depicted an elderly woman sobbing as she spoke directly to the camera. While very convincing, there are some inconsistencies around the woman’s mouth as she speaks, and she has an unnatural glow in her eyes.

RTVE fact-checkers got a high probability rating by running the video through an AI detector site. The footage could be traced back to a TikTok page run by a person who generates a lot of similar footage about news events. VRT reports that on his Facebook page, the creator even has a video showing his process for prompting ChatGPT to create materials.

Making sense of war with maps

When information is scarce, valuable insights can often be derived from open sources like satellite imagery.

That was what BBC Verify found when its team analysed the recent events for its analysis on how the RSF carried out a massacre.

After almost two years of holding el-Fasher under siege, satellite images showed that RSF troops had begun to construct a raised sand barrier around the perimeter of the city, blocking aid and sealing access routes in and out of the city, Merlyn Thomas, Matt Murphy, Peter Mwai & Richard Irvine-Brown write. “By early October the ring completely surrounded the city - with a smaller barricade encircling a neighbouring village.”

The BBC also verified several authentic videos filmed in late October, when the RSF overwhelmed el-Fasher and seized the city. “Soldiers were filmed laughing as they toured the abandoned headquarters carrying a grenade launcher. Later that day, RSF commander Abdul Rahim Dagalo - brother of RSF chief Mohammad ‘Hemedti’ Dagalo - was seen inspecting the base.”

Another graphic video analysed by the BBC Verify team showed the aftermath of a massacre at a university building in western el-Fasher, where dozens of dead bodies were seen scattered across the floor.

Meanwhile, as news spread of the Yale researchers’ findings about atrocities visible in satellite images, social media users began zooming into Sudan on Google Maps, claiming to have found more clusters and stains across the countryside. A TikTok video about the maps had nearly 10 million views just a few days after it was uploaded.

However, a closer inspection of the satellite imagery being used in Google Maps showed that it dated back to March 2024 — “Too old to show the current atrocities,” write VRT nws check’s Debackere and Nicolay.

On Google Earth, it’s possible to scroll back through older satellite captures. This confirmed that the same discolouration was there in previous years, even before the war broke out. “The figures seen in a row also resemble animals. This could be a cattle slaughterhouse, but that’s impossible to say for sure. This video certainly doesn’t show any recent atrocities by the RSF,” they write.

On November 6, there was hope for peace when the RSF said it had agreed to a truce. However, explosions were heard a day later near the army-controlled Sudanese capital Khartoum on Friday, according to AFP.

After a week, the fragile ceasefire was mostly holding — despite reports of drone and tank activity and several deaths of Palestinians as they began moving, once again, around the Gaza Strip.

The world was hoping the guns might soon fall silent for good in the troubled region, although much remained unclear about the two parties’ dedication to a lasting peace.

And if the arrangement does go to plan, the closing of one chapter opened another full of questions. Who will run the new Gaza Strip if not Hamas? Who will rebuild the now-unrecognisable enclave? When will enough aid arrive? And how will Palestinians be represented in the new governance structure that’s being planned by the West?

With much still unclear, an information void opened up, leading to the proliferation of theories across social media and in political circles. Reports emerged of Hamas punishing rival factions in the territory, while some Israeli ministers threatened a resumption of violence amid a row over the return of hostages’ bodies.

Fact-checkers and verification journalists have been studying these issues closely. Here’s our fact-check dispatch.

Hamas executes rivals

With Israeli forces withdrawing from parts of the Gaza Strip under the ceasefire deal, Hamas was working to reassert control over areas of the enclave — despite a stipulation in the agreement that Hamas must eventually disarm.

The Hamas crackdown included public executions of its internal enemies, groups and gangs of varying ideologies, some of whom are also seeking to govern over the Palestinians living in Gaza.

Videos from the most recent execution emerged earlier this week, as well as other footage showing displays of force. The footage was shared by accounts associated with Hamas, and Reuters reported that it was confirmed as authentic by a Hamas source. Hamas has pledged to crack down on “collaborators” and “lawlessness” in the Gaza Strip.

BBC Verify’s Merlyn Thomas, Matt Murphy, Benedict Garman, and Shayan Sardarizadeh investigated the incidents, and heard from analysts who said the “show of force is at least partially intended to counter armed clans which have increasingly challenged the group’s hold on the Strip over the past two years”.

Deutschlandradio’s Sarah Schmidt further narrowed down the location where the execution footage was filmed, via a source inside Gaza and cross comparison between the footage and satellite imagery, as well as comparison of street names which are inconsistent across different mapping software.

These incidents in the Gaza Strip have obviously come to the attention of U.S. President Donald Trump, who wrote on Truth Social on October 16: “If Hamas continues to kill people in Gaza, which was not the Deal, we will have no choice but to go in and kill them. Thank you for your attention to this matter!”

A previous execution in September also targeted members of a rival gang and led to graphic footage circulating of three blindfolded men being shot in a central location in Gaza City. The video and circumstances around it were analysed by Eurovision News Spotlight’s Alice Palussiere (France Télévisions), Sarah Schmidt (Deutschlandradio), Max Gilbert (BR24), Satu Helin (YLE), Valentine Joubin (franceinfo), and Maria Flannery (EBU).

Given Trump’s reaction, the ongoing internal struggle in the Gaza Strip, and external powers plotting a new course for the governance of the enclave, this story will continue to be an important one for journalists to monitor.

Disinformation surges after ceasefire announced

The early hours and days of the ceasefire saw much celebration in both Israel and the Palestinian territories of Gaza and the West Bank. However, some accounts on social media immediately began sharing misleading claims about the story.

While the remaining live hostages were returned from Hamas captivity, a series of posts circulated widely on X regarding the fact that none of them were female. One said: “All of the female hostages are dead. Where is the outrage? Savages”, while another said: “No female hostage is still alive. Think about that for a moment.”

These posts danced around a claim that all the women being held hostage were killed. However, the truth was more nuanced, as Tetyana Klug and Josh Axelrod reported for Deutsche Welle and Javier Menasalvas reported for VerificaRTVE.

While all the remaining living hostages who were returned to Israel in the last week were indeed men, most of the women had already been returned in previous exchanges, the majority in November 2023. There was only one female hostage still unaccounted for earlier this week, Inbar Hayman, who was known to be dead. Her remains were handed over later in the week.

There were other claims, including images that were old or taken completely out of context. For VRT nws check, Daan Nicolay investigated a viral image of a Teletubbies doll that appeared to have been booby-trapped, with wires seen inside the soft toy.

According to the accompanying post, the Gaza Civil Defence had found innocent items that had been tampered with by the Israeli army. However, a reverse image search revealed a YouTube result as old as 2018, and the video is probably from Yemen.

According to old news reports from the time, it was the work of Houthi rebels, VRT reported, although it was not possible to confirm the exact origin or context surrounding the video, due to the fact that it was published among pro-Saudi quarters, which opposes the Houthis.

In another case, a video circulated that appeared to show a mobilisation of Hamas on the streets of Gaza, with a convoy of cars and militants wearing the typical uniform of black with green headbands. However, this was also old, Eurovision News Spotlight member fact-checkers found.

DW’s Tetyana Klug and Josh Axelrod said: “Some users in the comments interpret the image as being taken after the current ceasefire went into effect. That is not what this is.” According to RTVE’s Javier Menasalvas: “This video is not current. It is footage filmed in January during a prisoner exchange in Gaza. A reverse image search located the footage in a January 25 Facebook post.”

The task to rebuild a devastated Strip

Amid all the political wrangling in this story, images from inside the Gaza Strip after two years of relentless bombardment brought up a simple question: How will the local population rebuild their homes and neighbourhoods?

Much of the enclave is an unrecognisable sea of concrete rubble — which is partially why it has been so difficult to verify and geolocate footage emerging from this vastly changed landscape.

So how big is the task of rebuilding? Our colleagues at BBC Verify probed what this might look like by speaking to experts. The UN estimates that the damage is worth approximately $70 billion, and according to Prof Andreas Krieg, a Middle East security expert, “it’s worse than starting from scratch” due to the sheer amount of rubble.

Jaco Cilliers, UN Development Programme special representative for the Palestinians, said the level of destruction in the enclave is “now in the region of 84%”, according to the report by Paul Brown, Kayleen Devlin, Erwan Rivault and Barbara Metzler.

As well as the built environment, the BBC Verify team investigated the dire situation regarding clean water, sewerage access and electricity supply in the territory. Much of the vital infrastructure of the society has also been damaged, and even before the war began, the power supply was unreliable and much of it was fed in from Israel.

The issue could well become political when world leaders involved in the rebuilding efforts have to decide how best to get essential services back up and running, and whether Israel should be involved as it has in the past.

For now, Palestinians will be focused on mourning the almost 68,000 dead and accommodating the more than 170,000 wounded as they try to return to something resembling their lives before.

But even when Gaza is rebuilt, the environmental toll of both the heavy bombardment and the necessary reconstruction work — and the personal toll for those who have survived — has yet to be counted.

The technology's merits have long been debated, but this year's sweltering summer created fertile ground for a new wave of misinformation.

As Spain battled devastating wildfires, conspiracy theories about the cause of the blazes took hold, often fuelled by old or AI-generated images mixed in with real footage.

The spread of false claims wasn't limited to the climate crisis.

Amid the chaos, a separate political assertion gained traction: Donald Trump's claim that he ended a war every month while in office. While experts give him credit for contributing to a few ceasefires, the sweeping, opaque claim required more scrutiny.

Here’s our Fact-Check Dispatch.

Climate crisis and the A/C debate

The climbing summer temperatures each summer usually give rise to a wave of discourse about climate change—not least whether the mercury really is rising at all.

Despite evidence that several countries are breaching heat records year-on-year, claims are still made that it used to be just as hot in the past. Is this accurate, or is our memory deceiving us?

Reporting for BR24’s #Faktenfuchs, Sophie Rohrmeier and Claudia Kohler dug into the figures for Bavaria, which showed that the number of hot days have increased since 1951.

Some dismiss the facts of the warming planet as “scaremongering”, including the AfD’s Beatrix von Storch, who wrote on X: “We used to call it 'summer.' We weren't afraid of heat deaths then, and we didn't talk about heat protection plans. We got a day off from work, were happy, and went to the outdoor pool.” Her party colleague Stephan Protschka, who denies man-made climate change, was also vocal on the topic, saying reports of unprecedented heat every year were “absurd”.

Rohrmeier and Kohler spoke to a climate expert who confirmed “summer is changing”, and outlined how the cherry-picking of individual hot summers from the past are not an indication of what every summer was like in years gone by.

As record-breaking heatwaves swept across Europe this summer, a familiar debate re-emerged: the continent’s limited use of air conditioning compared to North America.

Every year, the heat plays a role in some 175,000 deaths in the EU, according to the World Health Organization. With temperatures in southern Europe regularly creeping towards 40 degrees Celsius, and in some cases surpassing it, is it time for us to finally embrace A/C?

There are several arguments against the technology in Europe, including a belief in some countries that the cold air makes you sick. But with temperatures rising as a result of climate change, prompted in part by our high energy usage, the carbon footprint of air conditioning units came into focus this summer.

Writing for France Info’s Vrai ou Faux, Thomas Pontillon reported on comments by National Rally MP Yoann Gillet, who stated that “no scientific study shows that air conditioning is harmful to the environment. None.” His comments came as the party was calling for a major plan to install air conditioning in buildings like schools, hospitals, and workplaces.

But, as Pontillon writes, air conditioning is harmful to the environment, and there are several studies to back this up. “The Ecological Transition Agency states in its latest estimate that air conditioning in France emitted almost 4.5 million tonnes of CO2 in 2020. This is almost 1% of the country's greenhouse gas emissions that year. This may not seem like much, but in 2020, only a quarter of homes and less than half of businesses were equipped with air conditioners,” he outlines. Firstly, the appliances are powered by electricity, but an even greater polluting factor is the refrigerant gas that is used in air conditioners—some of which escapes from the machines and emits significant CO2.

With any luck, these appliances and the refrigerants used in them will become more environmentally friendly in the years to come. For now, governments, businesses and homeowners will have to balance need against emissions.

Spain’s devastating wildfires spark theories

With the high temperatures also came devastating fires in several countries, with Spain and Portugal worst hit. More than half a million hectares have been burned in Europe so far this year, in one of the worst waves of wildfires to hit the region in years.

Thousands of people were forced to evacuate their homes in Spain and several were killed. The news story brought genuine, frightening images to our screens. However, the interest in the story also prompted a deluge of old or false images claiming to be related to the fires, as well as unsubstantiated claims about the disaster.

For example, a teenager was arrested for allegedly starting several fires in Galicia, and social media users immediately jumped to the conclusion that he was a foreigner. VerificaRTVE reported that the local authorities confirmed the person had no foreign nationality. In another case, an image circulated showing firefighters covered in soot—which was linked to Australia in 2019, even though this newer version bore some signs of having been digitally manipulated. The same image circulated in 2024 when there were fires in Argentina.

There were also plenty of AI-generated images in circulation as the fires took hold. VerificaRTVE reported on several dramatic shots claiming to show the fires in Castile and León and of firefighters resting on the ground as the flames rose up around them.

Using a range of AI detection tools and reverse image searches, the reporters confirmed the provenance of the Castile and León image. They also managed to find it on Facebook, where it appeared with an AI-generated label.

Besides visual fact-checks, the verification team at Spain’s national broadcaster also published a number of explainers giving the facts on the way forward, the law, and the conspiracy theories that spread as prolifically as the fires themselves.

Speculation on the summer’s fires involved claims that they were located near rare earth deposits, and were set alight with an aim of extracting the materials. There were also claims that the land was destined for wind farms or solar power plants. In actual fact, any forested areas that are destroyed are generally required to be restored by the regional authorities.

Amid the discourse, there were also questions about whether enough was being done to prevent these disasters. VerificaRTVE broke down the responsibilities held by local, regional and national authorities, which each have mechanisms and laws they must follow. There are also helpful methods being used by some communities, such as preventative agricultural grazing. Educational articles like this one are important civic resources, and enable people to monitor whether enough is being done in their area.

Trump claims fact-checked

In August, Donald Trump deployed the National Guard in Washington, D.C. and placed the city’s police department under federal control, declaring that crime in the U.S. capital was “out of control”. It sparked concern that the president may use this as a blueprint to seize control of other American cities where there is strong dissent to his leadership.

But how true were the claims of D.C.’s danger, really? Reporting for Deutsche Welle, Astrid Prange de Oliveira writes that the claim that the District of Columbia is “among the most violent jurisdictions in the United States” was misleading. D.C. is not a state, so a comparison with other major cities is more appropriate, she writes. And as to how it stacks up against other cities: homicide rates were higher in St. Louis, Missouri, and New Orleans.

Another of Donald Trump’s high-profile claims also came under scrutiny—in which he said that he had settled six wars in six months. SVT Verifierar’s Henrik Silver, Linus Svensson, Stephanie Mourad, and Simon Krona probed the president’s efforts as a peace broker. Speaking to a peace and conflict researcher, they found that the claim was exaggerated.

White House press secretary Karoline Leavitt outlined what the six wars were: Thailand and Cambodia, Israel and Iran, Rwanda and the Democratic Republic of Congo, India and Pakistan, Serbia and Kosovo, and Egypt and Ethiopia. In another press conference, Leavitt referenced “the closing of seven global conflicts”.

Professor of peace and conflict research at Uppsala University, Isak Svensson, gave Trump credit for his contributions to ending some of these conflicts, but even in those cases, not all represented a sustainable peace.

Armêl Balogog writes for France Info that the U.S. president has played an important role in bringing peace to three conflicts since he returned to power. A peace agreement signed between Armenia and Azerbaijan in early August ended four years of conflict. In Cambodia and Thailand, a ceasefire agreement ended a five-day war in July. And in Israel and Iran, calm was restored after a 12-day war in June, although as Balogog notes: “It should be remembered that the United States also participated in that war by bombing Iranian nuclear sites”.

“Some experts also give Donald Trump credit in a fourth conflict, the one between India and Pakistan last spring (…) The American president welcomed having obtained a ‘total and immediate ceasefire’ in this conflict. But India denies any involvement of the United States in the peace negotiations. According to it, there was no mediation; the conflict was settled between the two belligerents.”

You may notice that one of these—Armenia and Azerbaijan—was not mentioned on Leavitt’s original list. Balogog writes: “To reach six conflicts, two are missing. However, the American president doesn't always mention the same ones, and neither does the White House.”

This is why wading through the web of claims coming from the White House is not just difficult, but a crucial act of journalistic due diligence, especially in an age when readers are more likely to share a headline than read the story it represents.

Claims circulated by AI chatbots like Grok and pro-Israel social media accounts have falsely asserted that photos of severely malnourished children are either old, from other conflicts, or that their condition is due to pre-existing medical conditions rather than starvation.

Against a backdrop of increasingly dire warnings from the World Health Organisation and numerous aid organisations, Israeli officials have disputed the existence of a famine and blamed logistical issues on UN agencies and Hamas.

Political figures have also contributed to the confusion, with false statements about the lack of European aid to the region being quickly debunked. As the information war intensifies, members of Eurovision News Spotlight continue to provide essential context and clarity to a complex and politicised crisis.

Here is our Fact-Check Dispatch.

Grok gets it wrong on Gaza's malnourished children

Since its release to users, Elon Musk's AI chatbot Grok has been widely criticised for repeatedly spreading misinformation and generating hateful content. Reports from mid-2025 detail how the bot made antisemitic remarks, including linking Jewish surnames to "anti-white hate," and even praised Adolf Hitler in response to user prompts, at one point referring to itself as "MechaHitler."

More recently, Grok has been used to spread false information about images of malnourished children in Gaza.

Reporting for Deutsche Welle, Monir Ghaedi and Shubhangi Derhgawen found that Grok has repeatedly provided misleading or false claims about their origin.

Ghaedi and Derhgawen examined a response made by the chatbot after a user asked Grok for verification assistance of a photo purporting to show a young girl begging for food in Gaza.

While the image itself has been confirmed as having been shot in Gaza City on July 26 by an Associated Press photographer, Grok wrongly identified it as having been shot in Iraq in 2014.

"This photo was taken on August 10, 2014, near Sinjar mountain in Iraq, showing a Yazidi girl fleeing ISIS violence on a truck. It's not from Gaza in 2025, despite recent misattributions," Grok responded, which gave rise to a flood of users echoing Grok's assessment.

A report by Paula Peña for VerificaRTVE also confirmed the authenticity of the photo of the young girl, captured at a food distribution point in Gaza City, debunking claims that it was an AI-generated image or old content from Iraq or Syria.

Similarly, a campaign by pro-Israel social media accounts attempted to sow doubt about photos of a two-year-old boy named Yazan and another child on his mother's lap, claiming they were old images from the 2016 famine in Yemen.

Nele Baeyens, Ellen Debackere and Bram Vandendriessche of VRT NWS investigated the claim, concluding the photos are indeed recent and originate from Gaza.

apnews

A post shared by @apnews

"The accounts disputing the images are wrong in this case. A simple Google search reveals that the photo shows a two-year-old boy. His name is Yazan and he was photographed on Wednesday, July 23, in the Al-Shati refugee camp in Gaza by Omar Al-Qattaa for the AFP news agency," they reported.

afpphoto

A post shared by @afpphoto

Aid groups contradict Israeli claims on famine

The debate over whether a famine is occurring in Gaza has become a central point of contention between Israeli officials and aid groups.

While the Integrated Food Security Phase Classification (IPC) reported that "the worst-case scenario" of famine is unfolding, with at least 16 children under five having died from malnutrition since mid-July, a coordinated disinformation campaign has been used to downplay the crisis.

Jenny Hauser of Eurovision News Spotlight examined claims by Israeli government spokesman David Mencer that Hamas and aid agencies are to be blamed for what the World Health Organisation has described as "man-made mass starvation" in Gaza caused by Israel's blockade.

Speaking to Sky News in July, Mencer said, "Israel causes no famine. There is a man-made shortage, but it's been engineered by Hamas. That's the point. That's the end of the sentence, which you don't include. But this suffering exists because Hamas made it."

The UN and aid agencies have disputed these claims, Hauser writes.

During a press conference on July 23, Head of the World Health Organisation, Tedros Adhanom Ghebreyesus, told reporters: "I don't know what you would call it other than mass starvation, and it's man-made, and that's very clear.”

VRT NWS' Baeyens, Debackere, and Vandendriessche further examined the coordinated narrative by Israeli officials, which included posts by the Israel Army, Israeli television channel ILTV Israel News and pro-Israel supporters, all of which echoed the government narrative that there is “no starvation in Gaza”.

"On social media, people try to reinforce these messages by claiming that images of very thin children are either outdated or that illnesses are causing them to be so thin. Pro-Israel influencers are spreading this message through various channels," Baeyens, Debackere and Vandendriessche report.

"There is insufficient evidence to determine the children's precise medical condition, but according to medical experts, it is likely a combination of illness and malnutrition that is causing their thinness," they said.

Trump's claim that "no European country" has given humanitarian aid to Gaza

During a press conference on July 27, where President Donald Trump announced a trade agreement that would impose 15% tariffs on European imports into the United States, the president was asked a different question: “Should Israel do more to allow food into Gaza?”

"Well, you know, we gave $60 Million dollars two weeks ago, and nobody even acknowledged it for food. And it's terrible. You know, you really at least want to have somebody say thank you. No other country gave anything. We gave $60 million two weeks ago for food for Gaza. And nobody acknowledged it. Nobody talks about it. And it makes you feel a little bad when you do that. And, you know, you have other countries not giving anything. None of the European countries, by the way, gave. I mean, nobody gave but us and nobody said, gee, thank you very much. And it would be nice to have at least a thank you."

Romane Bonnemé and Grégoire Ryckmans of RTBF concluded that Trump's statement was false.

"The European Commission announced in January 2025 that it would release €120 million as part of a new humanitarian aid program for the Palestinian enclave. A few months later, on May 21, 2025, the European executive confirmed the allocation of an additional €50 million, bringing the total aid distributed to Gaza to at least €170 million in 2025 and to more than €450 million since 2023," Bonnemé and Ryckmans reported.

“Furthermore, as of June 12, 2025, " 74 flights had arrived in Gaza and the EU had transported more than 4,800 tons of goods to Gaza (by air, sea or road)," the Council of the European Union (EU) website reported . These essential goods included " medicines, medical supplies, shelter, water, sanitation and hygiene items, and educational items on behalf of humanitarian partners, as well as stockpiled items belonging to the EU," they said.

SOURCES

Monir Ghaedi and Shubhangi Derhgawen (2025). Available at: Fact check: Grok gets it wrong on Gaza’s malnourished kids. [online] dw.com. [Accessed 5 Aug. 2025].

VRT NWS (2025). Available at: Foto’s van ondervoede kinderen in Gaza leiden tot informatieoorlog op sociale media: wat is er aan de hand? | VRT NWS: nieuws. [online] VRTNWS. [Accessed 5 Aug. 2025].

Hauser, J. (2025). Available at: ‘There is no famine in Gaza, there is a famine of the truth’: Fact-checking Israeli government spokesperson David Mencer. [online] Spotlight.ebu.ch. [Accessed 5 Aug. 2025].

‌Par (2025). Available at: ‘Aucun pays européen’ n’a donné d’aide humanitaire à Gaza : cette déclaration de Trump est fausse - RTBF Actus. [online] RTBF. Available at: [Accessed 5 Aug. 2025].

Peña, P. (2025). Available at: Esta imagen de una niña en un reparto de comida es real: está tomada en Gaza el 26 de julio. RTVE.es. [online]

As Pride Month closed, there was a resurgence in LGBT-related misinformation, and an examination on the long-running conspiracy theories facing several famous women who are accused — without evidence — that they are secretly transgender.

Confusion reigned over a claim from the Gaza Humanitarian Foundation that it does not hand out sugar, contrary to previous evidence, while in the UK, legal experts unpicked the likelihood of a case being brought against punk duo Bob Vylan for its “Death to the IDF” chants.

Here’s our fact-check dispatch.

AI & old content stokes anti-migrant sentiment

Anti-immigrant rage-bait content is nothing new on European social media. Even more common are the speculative narratives about the proliferation of Islam, which have been widespread for at least a decade.

What is newer about these types of narratives is the vehicle being used to spread them. Where out of context clips were once shared to exaggerate the presence of certain ethnic groups in European cities, artificial intelligence is now able to generate the exact scenes sought by the spreaders of these theories.

The propaganda issue could become even worse, reports Bram Vandendriessche for VRT nws check. Google’s new Veo3 model, which turns text to video, can create very realistic videos and has just been launched in Belgium, he writes.

Among the AI-generated clips Vandendriessche analysed were interviews with asylum seekers coming ashore in the UK, who talked about beautiful women and free houses and money from the government. Another example was a series of clips about cities like Brussels and Madrid and how they might look in 2050. These videos included racist depictions of Muslims with the streets full of rubbish and camels. Meanwhile, eastern European cities like Moscow and Budapest “look neat and prosperous”, Vandendriessche writes.

“The rather cartoonish videos about the future are not intended to come across as ‘real’ but are mainly intended to provoke emotional reactions and thus gain more likes,” he adds. The videos show that even when AI-generated videos are not being presented as real, they are still being leveraged to influence political opinions and ride the rising wave of anti-migrant sentiment across Europe.

There were plenty of examples of similar content being shared around amid the heat wave that baked the continent over the past week. For Eurovision News Spotlight, Jenny Hauser unpacked a viral video shared by serial misinformation spreader @RadioGenoa, which shares almost exclusively anti-immigrant content.

The probe found the video was at least partially old, as one of the clips in it was found in use in 2023. The footage seemed to be shared to suggest immigrants were ‘flooding’ Paris, but as Hauser notes, this is already the most ethnically diverse part of France.

How AI fuels transphobia & Pride Month debunks

The list of famous women who have found themselves at the centre of false claims about their gender is long. For many of them, like Brigitte Macron, Imane Khelif, and Michelle Obama, the theories have endured over years, resurfacing every few months when they are back in the news.

This genre of misinformation is, once again, nothing new. When Lady Gaga first burst onto the music scene in the late 2000s, for example, she was the subject of a years-long misinformation campaign claiming that she was intersex, which was even covered by mainstream media outlets at the time.

gaytimes

A post shared by @gaytimes

For VerificaRTVE, Paula Peña tracked the newer prevalence of AI-altered images that push these narratives forward. Old pictures of famous women, changed with AI to make them look more masculine, are widely shared and reshared with claims that they are transgender. The false images purport to show proof that they were once men.

The wife of the French president, Brigitte Macron, is a very common target for theories about her gender identity. In fact, we have already reviewed a long-running conspiracy theory about her for Eurovision News Spotlight.

In recent days, the discourse took a new turn, as Armêl Balogog reports for franceinfo’s Vrai ou Faux. An article claimed that a surgeon who had proof of Brigitte Macron’s transgender identity died after a fall — suggesting foul play.

The investigation by Vrai ou Faux found the website hosting the article was impersonating five real journalists to lend its stories an air of credibility. “This process is reminiscent of a well-known method used in pro-Russian interference operations, which involves hijacking the identities of media outlets to spread fake news, except this time they are targeting individuals,” Balogog writes.

Artificial intelligence was used to create a deepfake video featured in the article. Journalists who do exist, and whose names are credited as authors on the website, are considering legal action, Balogog reports.

Elsewhere during Pride Month, an image circulated showing the inside of a metro car decorated with the rainbow Pride colours, claiming to show a subway in Spain. As Javier Menasalvas from VerificaRTVE reports, it is actually from Rome in 2024.

There was also a video showing Donald Trump announcing that he was cancelling the celebratory month-long festival, which turned out to be an AI-altered video, reports VerificaRTVE.

Top magazines invent fake writers for AI-generated articles

An investigation by VRT nws check’s Tom Buytaert , Bram Vandendriessche, and Daan Nicolay revealed an extraordinary story for the media industry: the Belgian edition of Elle magazine created fake journalists and enlisted the help of AI to write more than half of its online output across several months, without disclosing it to readers.

The investigation further found that other brands owned by the same parent company, Ventures Media, were doing the same, including Marie Claire and Forbes. The company said it was a test, but after questions from the VRT journalists, they changed the profiles of the article authors and added a disclaimer to the articles “written” by them to say they were actually generated using AI.

There were a number of factors that served to tip off the VRT journalists. For one, photos used for the authors were found on a database of AI-generated headshots. Through the Wayback Machine, they were able to find that one of the authors even underwent a name change but kept the same image. Her email address as listed on the website did not exist.

The investigation even uncovered a psychologist from Psychologies magazine who did not actually exist, despite references in her articles to “my practice”. Carl Defreyne, chairman of the Psychologists' Commission, told VRT that psychologist was a protected title and added: “The example of Femke, who presents herself as a psychologist under a journalistic profile and provides advice from that role, is not only misleading, but also legally and socially unacceptable.”

Gaza Humanitarian Foundation claims it doesn’t hand out sugar

The Israel- and U.S.-backed Gaza Humanitarian Foundation has begun posting on X about what it says are false claims made about events at its distribution centres and the aid it hands out.

On July 1, the GHF published a screenshot of a post by a Palestinian man claiming personnel at a GHF distribution site had dug a trap for people trying to get to sugar — and that it said had been placed in a separate area from the remainder of the aid.

In a bizarre development, the aid foundation said the claim could not be true because it “is not distributing sugar”, Jenny Hauser writes for Eurovision News Spotlight. The head of the organisation, Rev. Johnnie Moore, doubled down with the same statement.

There were no widespread claims about the “trap” alleged by the original poster. However, Hauser looked into the statement that the Gaza Humanitarian Foundation does not hand out sugar at all. “There are a number of reports appearing to refute this claim. An Associated Press report from May 30 cites a GHF spokesperson that the food boxes being handed out contain sugar among other food products. Photos in news reports by the Middle East Eye and ABC News of the content of boxes handed out by the GHF also showed packages with the Hebrew word for sugar on them. Eyewitness footage and photos shared on social media also showed sugar inside the aid boxes,” Hauser writes.

In fact, the below is an image published directly by the Israeli military on May 27 along with a statement about the establishment of GHF sites, and a caption: “Photos of humanitarian aid packages”. Packets of white sugar are clearly visible.

Legal experts weigh in on Bob Vylan chants

The controversy that beset the Glastonbury music festival in the UK lasted throughout the whole week after police said they were investigating a performance by punk rap duo Bob Vylan.

The group’s singer led chants of “Death, death to the IDF” and “Free, free Palestine” during the set on June 28, which was broadcast live on the BBC. In the UK, some politicians and commentators said the chant amounted to antisemitism, and Prime Minister Keir Starmer labelled it “hate speech”. Confirming an investigation, the police said they were probing potential hate crime or public order offences.

Even before any police investigation had been carried out, Bob Vylan was dropped by its management and agency. The group has denied antisemitism and pledged to continue to speak up for Palestinians.

bobbyvylan

A post shared by @bobbyvylan

For RTÉ Clarity, Jack McCarron and Kate McDonald asked British legal experts if there was a clear legal basis for any case against the duo. Professor David Mead told them the prosecution would need to be satisfied that they have enough evidence to ensure a realistic prospect of conviction, and would need to decide whether bringing charges was in the public interest. Meanwhile, Jonathan Hall KC said a section of the Public Order Act relating to “stirring up racial hatred” could be used, but added police could struggle to bring a charge.

“It's quite hard to show off the back of saying ‘Death to the IDF’ that he intended people in the audience to hate Jews … that's because what he said was to a military of a country — although, there is an exceptionally strong link between Jews and that country,” he said.

Both experts agreed that the legal test hinges on how to interpret the intent behind the chant. As this story shows, leaning on the expertise of specialists can be an invaluable way to reach a fuller analysis on a complex story.

Since the war began, members of the Eurovision News Spotlight network have been working tirelessly behind the scenes, sifting through the deluge of content from both sides to expose manipulation and verify the truth.

In this edition of the Fact-Check Dispatch, we see how quickly misleading content emerges, from repurposed old videos to sophisticated AI-generated deepfakes.

We'll delve into specific examples, revealing how readily old footage is recontextualised and how advanced AI tools are now weaponised to distort realities and inflame narratives, making the online landscape increasingly treacherous for anyone seeking accurate information.

Here’s our fact-check dispatch.

Operation Rising Lion

Due to limited internet access and a reliance on government and military sources for information about significant events in Iran, members of the Spotlight network heavily relied on geolocation techniques to verify the accuracy of their content. They also used these methods to explain the stories to their audiences.

Maria Flannery and Derek Bowler of Eurovision News Spotlight cross-matched IDF video and graphics with Google Earth and satellite imagery from Maxar Technologies to build an accurate map of the nuclear sites being targeted by Israel last week, and over the weekend, the United States.

Earlier in the week, Israel said it was targeting a number of above-ground nuclear facilities, including Natanz, Isfahan, Arak/Khondab, and a number of offices and manufacturing facilities in Tehran that Israel said were linked to the nuclear programme.

In the background, speculation grew about what it wanted to do about Fordo, a fortified uranium enrichment facility built deep within a mountain. Bowler and Flannery write: “Constructed in 2006, the Fordo facility is located approximately 80 metres (260 feet) underground. Officials stated that the decision to build the facility underground was made due to threats from Israel to conduct air strikes on such installations.”

The site was widely understood to be beyond the reach of any of the bombs in use by Israel, despite the fact that Netanyahu had claimed that his army was capable of hitting all of Iran’s nuclear facilities. The weapon that could reach it? American-made “bunker-buster” bombs — formally named GBU-57 Massive Ordnance Penetrators —the heaviest non-nuclear weapon in the U.S. arsenal at 13,600 kg.

Days after this analysis, this topic of much speculation did indeed come to pass, with the U.S. deploying its B-2 Spirit bombers to drop several “bunker buster” bombs on Fordo in a mission that also struck Isfahan and Natanz in the early hours of Sunday morning.

With the news that U.S. bombers had hit Fordo, BBC Verify’s Visual Journalism team published detailed maps showing the landscape around the underground facility and comparing the before and after satellite imagery that confirmed the strike sites and gave us more insight into how many bombs were dropped on it.

While speculation reigned in the West about whether or not the U.S. would get directly involved, other information was being released in Iran, as the France24 Observers write.

“Israeli drones successfully neutralised several of the Islamic Republic's anti-air defences deep inside its territory. What is particularly surprising is that some of these drones were manufactured in covert production facilities deep within Iran. Iranian state TV has broadcast images showing camouflaged lorries and vans that were used to transport the drones, as well as images of the makeshift FPV factories,” the team writes.

Such images would be impossible to independently verify as there is very little information to work with in terms of geolocation. To analyse them, the reporters spoke to an expert on Iranian weaponry, Farzin Nadimi, who said: “We have known for many years about the deep infiltration of Israeli assets, almost at every level of the IRGC, the IRGC intelligence service, other intelligence services of Iran, the military and among politicians.”

Nadimi told France24 that much of the Iranian intelligence services are trained for internal security risks, such as protests and enforcing domestic repression, which they see as a more urgent threat than a foreign attack. On the drones, he said it would not be too complicated to smuggle parts into Iran, but added: “For a complex mission like this, however, you can't rely entirely on locals and local assets, you need to have agents on the ground. They had planned this for a long time. They claim to have been working on this plan for many years.”

As Israel pounded Iranian facilities earlier in the week, SVT Verifierar tracked the damage inflicted on Isfahan, Natanz, and Kermanshah, which has been named as a base for ballistic missiles. Maria Lapenkova, Ellika Nilsson, and Simon Krona mapped the damage using satellite imagery, and tied together videos from the incidents with damage visible in Maxar imagery.

Misrepresented content distorting facts in real time

For those looking to discover, verify, and clear the rights to eyewitness media during a breaking news story, there is a golden window of about 30 minutes before the landscape becomes inundated with old, unrelated videos repurposed by users to appear as part of a story.

As social media matures and the use of AI-generated content accelerates, that window is shrinking rapidly. The ongoing war between Israel and Iran has significantly contributed to a surge of misrepresented content online, making it more challenging to separate the news from the noise.

Baghdad 2003

Staying with content maturity, Maria Flannery of Eurovision News Spotlight investigated a viral TikTok video posted with a claim it showed a heavy bombardment of Israel on June 17.

The video (archived here) was posted with onscreen text reading: “Iran Has Launched The Largest Attack On Israel In History. Today 17/06/25. Iran New Wave Of Missile Hitting Israel.”

@lahoreheraldnewsIran has launched the largest attack on Israel in history now today 17/06/2025 @skynews @lahoreherald #foryoupage❤️❤️ #NEWS #breakingnews #israel #war #Lahorehrald #lahoreherald #iran #telaviv #latestnews #today #warupdate #fyp #newvideo

Tiktok failed to load.

Enable 3rd party cookies or use another browser

A reverse image search on Google Lens of different stills from the video yielded numerous examples of it being shared in recent days.

“However, it also pulled up a near-match from a CNN link titled: ‘Iconic moments from 2003 Iraq War’. The video thumbnail looked very similar, but unfortunately the website’s video player was not working at the time of writing,” Flannery writes.

“After this, we carried out manual searches online to find archive material from the Iraq War. It led us to this — a video on the ITN Archive YouTube page titled: ‘Iraq War: Shock and Awe Assault on Baghdad Begins (2003)’,” she continues. Buildings, the general landscape and even the nighttime lighting of the areas struck matched scenes from the viral TikTok video.

“Another unusual aspect of the TikTok video is how it appears to feature the sounds of shocked eyewitnesses, even though we now know a news crew professionally filmed it in a war zone,” Flannery reported, adding that the audio — which did not sound like Arabic, Hebrew or Farsi — was likely added to the TikTok to make it appear more realistic.

The team at DW Fact Check also debunked Baghdad 2003 footage after analysing a one-minute compilation published on TikTok (archived here) described as showing multiple bombings from the air, explosions and fire at night in Iran.

@www.jimmykenyaISRAEL BOMBING IRAN WITH NO MERCY #FYP #foryoupage #israel #us #iran #iran🇮🇷 #israel🇮🇱

Tiktok failed to load.

Enable 3rd party cookies or use another browser

“A reverse image search of stills from the compilation reveals two sources that prove the time and place for sequences of the video. The first sequence of the TikTok video is part of a longer video compilation of a bombing at night in Iraq, published by the US media outlet CNN,” write Uta Steinwehr and Adnan Sidibe.

Itamar Ben Gvir and Omer Bar-Lev

Nora Litoussi of France 24 looked at a trending video showing a heated exchange between Israeli politicians, Itamar Ben Gvir and Omer Bar-Lev, with posters claiming that the two politicians were fighting after the Iranian strikes carried out on June 13, 2025.

However, Litoussi reports that the conversation took place in March 2022 and has no ties to the current conflict.

“We ran the video through a reverse image search (see our easy guide to find out how to do one yourself). We discovered by doing this that the same video was published online back in March 2022 on the X account of the Israeli TV channel, Kan – a full three years before the current conflict between Israel and Iran broke out. The video actually shows a confrontation between the two men that took place on March 27, 2022, in the Israeli city of Hadera after a deadly attack took place there,” Litoussi writes.

Staying at France 24, Nathan Gallo investigated online claims that several Chinese aircraft have landed in Iran to supply the country with weapons.

Screenshots shared online from the air traffic monitoring site Flightradar24 (archived here and here) showed that flights operated by Luxembourg cargo airline Cargolux on June 14, 15 and 16 arrived in Iran. However, so far, none of the Flightradar24 screenshots posted online since Friday show any aircraft heading for Iran, Gallo reports.

“When contacted by the FRANCE 24 team, the Flightradar24 website explained that the aircraft icons that appear to be flying over Iran are in fact only ‘estimation data’ shown ‘when a user clicks on a flight’, to give an indication of the direction in which the aircraft should be heading. ‘Estimated positions are noted by the black trail colour,’ Ian Petchenik, the company's communications director, told our editors,” Gallo writes.

AI-generated content intensifies online war narratives

Since Israel launched a widespread attack on Iranian installations and facilities on June 13, AI-generated content has continued to fuel misinformation and disinformation online. This surge isn't just about the volume of false narratives; it's about their sophistication and reach.

Advanced AI models are being used to create deepfakes of political leaders making inflammatory statements, generate fabricated news reports that mimic legitimate media outlets, and even produce realistic-looking images and videos of events that never occurred.

Lea Deseille of France Info investigated a video purported to show a convoy of military vehicles (archived here) carrying dozens of Iranian missiles from a mountainous cave.

Posted by a pro-Iranian account, the video has garnered over 2.1 million views since it was posted on X on June 18.

“Upon closer inspection, we quickly notice that the mountains and rocks at the bottom right are moving in an unnatural manner. Everything suggests that the scene is not real,” Deseille writes.

Similarly, the team at VerificaRTVE examined further video (archived here) which claimed to show Iranian forces mobilising with missiles.

The team found several inconsistencies throughout the 36-second-long video, including the merging of soldiers into one, soldiers randomly appearing and disappearing, and malformations in the hands of those in the video.

“We analysed the video with an advanced AI-generated content detection tool (Decopy.AI), and the platform concluded that it was generated with this technology with a 95.94% reliability in prediction,” VerificaRTVE writes.

Michael Schlegel and Fabian Dilger of BR24’s #Faktenfuchs looked at a video claiming to show destroyed aircraft at an Israeli airport (archived here) after an alleged strike, a video which the duo determined was AI-generated.

“The passenger windows of the front two planes appear to be floating in the air in places where, in the video, holes are actually torn into the planes' fuselages. Furthermore, the ambulances appear to be moving at the same speed. One of the ambulances even drives through the frame twice in the same position,” Schlegel and Dilger write.

“One thing is certain: the airport video does not show real footage from the war between Israel and Iran in June 2025,” the pair reported.

After the United States military struck nuclear sites inside Iran overnight into June 22, a video (archived here) purported to show the aftermath of “America destroying Iranian nuclear reactors”.

The video was debunked by Derek Bowler for Eurovision News Spotlight, who wrote: “A Google Lens search of the video revealed an earlier version from June 18, in which the creator describes the footage as “Big explosion in a city | ai video.”

The emergence of Google Veo into the misinformation landscape

One of the tools being used to spread instances of misinformation and disinformation online is Google Veo. The text-to-video generative AI model, developed by Google DeepMind, creates high-quality, realistic videos up to 8 seconds long based on user text prompts, and can also generate accompanying audio.

Across the Atlantic, Jeff Yates and Nicholas De Rosa of Radio Canada debunked a video that claimed to show Iranian missiles hitting a building in Tel Aviv (Archived here).

A cropped, low-resolution version of the video was shared by Maram Susli, also known as Syrian Girl, who claimed the footage showed a strike on Tel Aviv.

Yates and De Rosa found that the video was generated using Google Veo.

“Despite what this account claims, the video is fake. Other copies of this same video exist where the Veo logo can clearly be seen,” they report.

At Deutsche Welle, Kathrin Wesolowski and Adnan Sidibe sought answers on a viral TikTok video (Archived here) which claimed to show Israelis protesting against the war with Iran, chanting for peace.

“This video allegedly shows hundreds of Israelis protesting against Israel's ‘war on Iran’. People in the streets waving Israeli flags are shouting: ‘Stop the war. Iran, we are sorry. We want peace!’ The video has been viewed more than 4.9 million times and shared on several social media platforms like Facebook,” Wesolowski and Sidibe report.

Wesolowski and Sidibe note several key inconsistencies in the video, including mouth movements and limbs that disappear and reappear. They also note that the video bears the Google Veo watermark and is eight seconds long—the typical length of a Veo-generated video.

Soon after, Iran responded, first sending drones towards Israel which were intercepted by the Iron Dome air defence system. On Friday evening, the Iranian response heightened, with footage showing scenes of missile fire raining down on Israeli cities.

Over the weekend, the two countries traded attacks. Iran reported that at least 224 people were killed since the Israeli attacks began on Friday. Israel reported 24 deaths in the same period.

On June 12, at least 270 people were killed when an Air India passenger plane crashed shortly after takeoff from the western Indian city of Ahmedabad. Just one of the 242 people on board survived, and dozens others were killed on the ground as the plane struck a student accommodation building near the city’s airport.

Both of these major stories led to massive amounts of content being shared online, not all of which was trustworthy.

Here’s our fact-check dispatch.

Israel-Iran conflict escalates

With a huge amount of content widely circulating online about the Israel-Iran conflict, newsroom verification teams were working to verify real videos from the various incidents.

However, there were also fakes in circulation, showing just how carefully everything must be checked during a big news event like this.

For Radio France’s franceinfo, Thomas Pontillon and Valentine Joubin urged caution about false videos. They write: “Most of them date from last October when Tehran targeted Israel after the death of the leader of Hezbollah in Lebanon, a very close ally of Iran.”

Pontillon and Joubin also warned readers to beware of accounts that appeared to be official sources but were not. We have covered some of these before — verified accounts for “Mossad” and the “Iran Military” have no connection to the official organisations they purport to represent. Some of them clearly state in the account bio that they are commentary or parody accounts, but others require a closer inspection. In this case, as in many others, the paid verification programme on X has muddied the waters when it comes to recognising official sources on the platform.

To accommodate such a fast-moving story, the team at BBC Verify was running a live blog updating the public on its verification work as it happened on Friday, June 13.

Sebastian Vandermeersch found that a video originating in Syria over a year ago was found to be in circulation again, purporting to show an Iranian drone tangled in power cables. It wasn’t the firswt time this video was posted false claiming to be new; it also did the rounds in April 2024, during another escalation between Israel and Iran.

Paul Brown gave information about satellite imagery from Iran that showed signs of damage at the Natanz nuclear facility. He writes: “The image was captured by the US company Umbra Space using Synthetic Aperture Radar (SAR), rather than more conventional methods. The result is a high resolution, black and white image of the earth’s surface. The benefit of this technology is that it cuts through cloud cover, which can often obscure images captured by conventional satellites.”

Meanwhile Rob Corp cautioned that on Friday afternoon, the Iranian government said it had placed “temporary restrictions” on internet connections, which would limit the amount of material that was coming out of the country showing the situation.

For Deutsche Welle, Alima de Graaf and Tilman Wagner write about many of the viral videos in circulation were actually fake. In one instance, a video with hundreds of thousands of views on TikTok claiming to show Iranian sites hit on June 13 was actually AI-generated. How did they figure it out? Several odd details, including a distorted teddy bear that looked too clean for its surroundings, and statue-still firefighters that vanished into thin air.

In another case, an image on X that racked up 4.5 million views was posted with the caption: “IRAN HAS LAUNCHED MISSILES AT ISRAEL.” At the time of the post, there had been no confirmation of any missile strikes yet — Iran first responded with drones.

More probing further confirmed that the image wasn’t from that day. De Graaf and Wagner write: “A reverse image search reveals the photo in the post is from a military exercise in Iran several years ago. The image first appeared online in 2021. The photograph appears to have been taken in southern Iran during the 17th Iranian Revolutionary Guard Corps (IRGC) joint Great Prophet 17 exercise in December 2021, which included the launch of rockets and missiles as part of a five-day military drill.”

Derek Bowler looked at another video for Eurovision News Spotlight, purporting to show Israeli drones hitting targets in Iran. A reverse image search revealed a version of it posted in October 2024.

Bowler writes: “This particular piece of content has been misrepresented in several conflicts, mostly recently during Operation Sindoor, after India launched missile strikes on Pakistan on May 7, 2025.”

Turning away from fake content, the team at the France 24 Observers analysed the initial images released by the Israeli military. They geolocated an area seen in an IDF video as being about 10 kilometres from Borujerd in western Iran, but were unable to say that the site had been hit.

Air India crash sparks misinformation

An Air India Boeing 787-8 Dreamliner was en route to London when it crashed shortly after takeoff in Ahmedabad. Immediately after reports emerged of the crash, footage and images were shared by local people showing clouds of black smoke rising into the sky. In among the genuine videos that were circulating after the crash, some items immediately raised suspicion.

One of them was a clip purporting to show the moment of the crash as filmed from inside the aircraft. Maria Flannery writes for Eurovision News Spotlight that it was actually filmed during a plane crash in Nepal in early 2023, which killed everyone on board, including the man filming.

“We were also independently able to find news coverage of the man’s final moments in 2023 via a Google Lens reverse image search. In addition to this, it is also possible to debunk claims this showed an Air India flight by comparing colour scheme and branding of the interior of the plane.”

For VerificaRTVE, LM, Paula Peña and Mario Pérez write of another similar example showing smoke filling up the inside of a plane. This was actually filmed during an incident on a Ryanair plane that took off from Bucharest in 2020.

Both Eurovision News Spotlight and VerificaRTVE also uncovered a number of examples of AI-generated content being circulated, presented as photos of the crash.

Israel’s online ad campaign for Gaza tourism utopia

The state of Israel bought millions of euro worth of online ads targeting European countries including Finland, an investigation by YLE’s Satu Helin, Eero Mäntymaa, Samuli Huttunen, and Ville Juutilainen found.

One such video created using AI resembled a travel ad, showing happy tourists, beautiful beaches and luxury hotels and bars. The video claimed: This is what Gaza could have been if there were no Hamas.

“The video is an example of propaganda produced by the Israeli state, which Israel uses paid online advertising to spread,” the authors write. “The Eurovision ads, [a state-funded campaign for Israel’s Eurovision Song Contest entry] which also attracted attention in Finland, are only a small part of Israel’s long-running media campaign.”

With these ads, “Israel wants to convey its own message to the global audience,” the YLE team writes.

“In its advertisements, Israel has claimed, among other things, that there is no famine in Gaza, that Israel supplies humanitarian aid to Gaza without restrictions, and that the problems in the region are the fault of the terrorist organisation Hamas.”

The investigative reporters probed the contents of the ads and their reach across Europe.

In Gaza, the story of the week was the much-needed aid that finally made it in, and the chaos that surrounded some of its distribution. There, a swirl of misinformation surrounded a controversial aid delivery programme, and even official sources were spreading fake news about chaotic scenes at a flour-filled warehouse.

And the new pope, who has named AI as one of the biggest challenges to humanity, was himself ripped off in videos addressing world leaders.

Here’s our fact-check dispatch.

Conspiracy theorists milk Liverpool parade incident

A major incident at a victory parade for Liverpool FC in Merseyside, which saw a driver go through a crowd injuring more than 70 people, shocked Britain last week.

And at a time of heightened racial tensions and online right-wing discourse, any incident like this immediately prompts widespread speculation that the perpetrator was either an immigrant, a Muslim, an asylum seeker, or otherwise non-white and non-British.

The UK has been here before. After the Southport knife attack in 2024, which saw a teenager stabbing three children to death while they were attending a Taylor Swift-themed dance class, far-right activists flooded social media with misinformation which ultimately ended in riots, an attack on a local mosque, and missiles thrown at police.

In the case of Southport, it took several days for the identity of the attacker to be released. The flow of information by authorities was complicated by the fact that the attacker was himself a minor, limiting the amount of information that could be released before he was charged. The information vacuum in the days between the attack and the official naming of the suspect created the perfect environment for rampant conspiracy theories and the spreading of incorrect names, as reported widely at the time.

The Liverpool incident started out similarly. Right-wing activists immediately jumped on the incident as evidence of a criminal immigrant element in the community. But in an unusual turn of events, the British police released details about the suspect’s identity at unprecedented speed; he was a white British male aged 53, they announced within two hours of the incident.

The local police’s decision to release information early in a bid to stem misinformation was praised by the country’s independent watchdog on terror legislation, Jonathan Hall KC, who had previously warned that the failure to do the same in Southport had fanned the riots.

Nevertheless, misinformation still swirled in the hours after the Liverpool incident. For SVT Verifierar, Henrik Sköld and Simon Krona write about how posts circulated saying: “First image of Liverpool attacker. Attacker is thought to be foreign.” The account, @RadioEuropes, has since deleted the post, presumably after police released the information about the suspect. However, they later followed up with “Do you believe that the Liverpool attacker is a 53-year-old white British man?” and “Does Liverpool attacker look like a 53-year-old white British man?”.

In other posts, an image of a completely different man were circulating with captions like: “Young, and definitely not white. Coverup already running.” Sköld and Krona write: “On X, posts with this text have been viewed over a million times. But the man in the picture is not the one who drove the car into the crowds.”

The verification experts at SVT downloaded a moving clip from social media showing the wrongly accused man. Explaining how they analysed it, they write: “From that clip, we extracted several screenshots showing different facial expressions and angles of the man's face. Using facial recognition software, we were able to find a photo of a person resembling the man, published on a website belonging to a cleaning company outside Birmingham.”

Using Amazon Rekognition, the team compared faces and assessed the degree of similarity between them. “We were able to determine that the man who was wrongly accused was very likely the same person who appeared on the cleaning company's website … Based on this, we were able to identify the person, and assess that he did not look to be 53 years old, but rather much younger.”

As evidenced by the situation in Southport almost a year earlier, false IDs and race-based assumptions can have a hugely damaging effect on public safety, and can even spark riots. This further highlights the importance of both this type of verification work, and of a swift and transparent reaction by public authorities.

Misinformation swirls about Gaza aid crisis

The dire shortage of aid in the Gaza Strip came to a head during the week when the almost three-month-long Israeli blockade on goods finally ended.

A controversial U.S. and Israel-backed plan to deliver aid into the enclave via the Gaza Humanitarian Foundation failed immediately when desperate crowds rushed at the first distribution site it opened in southern Gaza.

Before that incident, it had not actually been clear whether the programme had begun distributing aid, after it said the day before that it had started handing out food. For Eurovision News Spotlight, Jenny Hauser looked at the evidence and spoke to people on the ground, ultimately confirming that the site was indeed operational near Rafah.

Meanwhile, the Israeli Ministry of Foreign Affairs claimed that Palestinians looted a Hamas warehouse storing flour, accusing the militant group of hoarding food. The official ministry account posted a video of the chaos at the warehouse, saying in the caption: “This is Hamas's warehouse in central Gaza — full of food, kept away from the civilians. People stormed in. Hamas opened fire.”

The Gaza government, meanwhile, said that the warehouse in Deir al-Balah was a United Nations World Food Programme property, adding that it was a “regrettable incident stemming from the catastrophic humanitarian conditions endured by the population due to the intensified Israeli siege and its systematic starvation policy”.

The World Food Programme confirmed it was operating the warehouse — not Hamas, as the Israeli Foreign Ministry claimed. The story of it being overrun by desperate people in search of food was true, Hauser writes.

The story about Gaza overlaps with a consumer story in the West, particularly for those in the pro-Palestine community who are keen to observe a boycott on Israeli goods.

However, it’s not always easy to know whether a product has come from Israel, as readers of VerificaRTVE pointed out in messages requesting an explainer. Paula Peña writes about a long-running theory that the number at the beginning of a product’s barcode indicates its origin, and that Israeli goods are barcoded with a number beginning 729. There were further claims that due to the boycott, Israel changed the number to 871.

Despite the detail in these claims, all is not as it seems, Peña writes. “This numbering does not indicate the origin of an item or its manufacturer, as explained by the [Spanish] Association of Manufacturers and Distributors (AECOC) in 2023. 871 is a code associated with the Netherlands, and Israel has not changed its barcode to avoid the boycott of its products due to the conflict in Gaza.”

In fact, the prefix identifies the GS1 member organisation to which the company is registered. The barcode prefix 729 is indeed associated with Israel, but does not mean an item was produced in that country or even that the company is based in Israel or of Israeli origin. Companies can choose to barcode their products through any of the GS1 member organisations, so it may just mean that the company registered through the Israeli member entity.

The clarification on what barcode numbers mean is valuable for both those observing boycotts as well as businesses that may be unfairly targeted by such actions.

Pope Leo targeted by misinformation

Pope Leo XIV is only a few weeks into his papacy, but even he hasn’t escaped being a target for widespread misinformation, including as the subject of AI-generated content.

For Deutsche Welle, Sarah Steffen and Kathrin Wesolowski looked into a number of examples, such as a claim that the pope responded to a letter by the Burkina Faso junta leader Ibrahim Traoré. The pope says in a video: “I have read your words not once, but many times (…), because in your voice I have heard not merely the anger of a president, but the righteous cry of a continent long wounded by the dual blades of neglect and exploitation.”

The 36-minute video is a complete fabrication. Steffen and Wesolowski found the real papal audience it drew from through a reverse image search. “The video used a morphing technique to make the Pope's lip movements line up with the AI-generated audio recording,” they write.

The video even drew comment from the Vatican’s official media portal. It’s unusual for the Vatican to put out a debunk article — this is the first under a new ‘Fake News’ tag on its website — but this is a new papacy. The newly inaugurated Pope Leo has made Artificial Intelligence one of his key priorities, identifying it as one of the big challenges facing humanity. He even referred to AI when he explained his choice of the name Leo. (The previous Pope Leo became known over 120 years ago for his defence of labour rights and criticism of free market capitalism amid the industrial age.)

Steffen and Wesolowski also explained how they fact-checked several other instances of false material, including a video purporting to show him criticising Donald Trump, which is manipulated, and his supposed ignoring of a Pride rainbow flag, which is actually an Italian peace flag.

There was also an unfortunate confusion about the very real and dire humanitarian situation in the Gaza Strip when a UN chief fumbled his figures on the child malnutrition crisis and severe lack of aid. The figures were later clarified, and still made for stark reading.

And despite the harsh restrictions placed on Russian shoppers since the war in Ukraine began, many of them have still been able to buy Western products, including luxury goods. Just how are they doing it?

Here’s our fact-check dispatch.

Trump’s false ‘evidence’ of white genocide in South Africa

A bilateral meeting between U.S. President Donald Trump and his South African counterpart Cyril Ramaphosa spiralled at the White House on May 21, after Trump claimed that white farmers in the African nation were being killed and “persecuted”.

Politicians and members of the press had gathered in the Oval Office on Wednesday for a meeting which was expected to ease tensions between the two countries, a week after Trump granted asylum to 54 Afrikaners who he claimed were discriminated against in the country.

As the meeting progressed, Trump surprised the room when he instructed a White House staffer to “Turn the lights down” in the Oval Office.

As the lights dimmed in the room, a TV displayed video footage of South African politician and the leader of the Economic Freedom Fighters party, Julius Malema, in which Malema is heard repeatedly saying to followers at a rally to “kill the farmer” and “kill the Boer.”

A second video on the screen, played from an X account, showed an aerial video of a road surrounded by agricultural land. The footage appeared to show dozens of vehicles stopped in the area, with white crosses on both sides of the road. Trump said they were “burial sites” of over 1,000 white farmers, and that the cars were mourners coming to pay respects at the graves.

President Ramaphosa said: “Have they told you where that is, Mr. President?, to which Trump replies “No”. Ramaphosa adds: “I’d like to know where that is..because this I’ve never seen.”

“I mean, it’s in South Africa,” Trump replies, to which Ramaphosa says, “We need to find out”.

Derek Bowler from Eurovision News Spotlight carried out a reverse image search on the aerial video that showed white crosses along a roadside. That brought up a news article from 2020, about a husband and wife who were murdered on their farm in Newcastle, South Africa, during a farm attack.

The white crosses did not represent “burial sites”, but were a temporary memorial to represent the lost lives of farmers after that 2020 double murder. The cars were there for a memorial procession for the couple.

At BR24’s #Faktenfuchs, Michael Schlegel writes that other images were shown in the White House that Donald Trump claimed showed “war crimes” against white farmers, including a printed out picture of body bags that had come from a news video. “The video shows footage from Reuters following a prison break in Congo in which more than 100 women were raped and burned alive,” Schlegel writes.

He also cites a quote by the video journalist who filmed the footage for Reuters, Djaffar Al Katanty, who told the news agency: “In view of all the world, President Trump used my image, used what I filmed in DRC to try to convince President Ramaphosa that in his country, white people are being killed by Black people.”

Another part of the video showed by Trump involved South African politician Julius Malema proclaiming the slogan: “Kill the boer, kill the farmer”. Astrid Prange de Oliveira from Deutsche Welle explains that the phrase comes from an Apartheid-era song, which has been classified as hate speech in South Africa on several occasions.

Malema was a member of the South African ruling party ANC, but was expelled in 2012 and later founded a more hard-left communist party, Economic Freedom Fighters (EFF). Ramaphosa has distanced himself from such statements, saying they are not government policy.

The story was also covered by María José Artuch at VerificaRTVE, who gave some background on the existing tensions between the U.S. and South Africa. “Trump signed an executive order cutting financial aid to South Africa and ordering refugee status and asylum in the US for Afrikaners,” she writes, adding: “Trump has also harshly criticised the South African government for its ‘aggressive’ stance toward the U.S. and its allies, citing as examples the complaint against Israel for genocide before the International Court of Justice and the strengthening of relations with Iran.”

Artuch writes: “South Africa has rejected these accusations and, in response to Trump's executive order, declared it ‘ironic’ that he would grant refugee status in the U.S. to a group that remains among ‘the most economically privileged’, while ‘deporting vulnerable people from other parts of the world’.”

For RTÉ Clarity, Laura Fletcher writes: “South African tech billionaire and ally and adviser to Donald Trump Elon Musk was present at the press conference in the Oval Office on Wednesday … it’s notable that many of the videos that made it into President Trump's montage were clips reposted by Mr Musk in recent months.”

Musk, who has been labelled by his own AI, Grok, as one of the worst spreaders of misinformation on his X platform, has railed against what he describes as the “genocide of white people in South Africa”.

He has also said his satellite internet service Starlink was “not allowed to operate in South Africa simply because I’m not Black”. The country’s communications authority said Starlink never submitted an application, but it is true that South African law requires investors to give local Black firms a 30% stake in their businesses, in a bid to tackle racial injustices of the past. Fletcher writes: “Such policies are a core pillar of the ANC after the end of Apartheid rule in 1994, but are very much at odds with the Trump administration's abhorrence of all things ‘diversity and inclusion’.”

At a time when every issue is so highly politicised, fact-checkers and journalists find themselves having to dive deep into almost everything the U.S. President says. In wielding such immense power, Trump is already making major changes to U.S. foreign policy, global geopolitics and economies around the world, some of which are based on false narratives such as this one.

Confusion over comment that ‘14,000 babies will die in 48 hours’

It was a major headline during the week: “UN says 14,000 babies could die in Gaza in next 48 hours.”

The figure came from chief of the United Nations Office for the Coordination of Humanitarian Affairs, Tom Fletcher, who told the BBC: “There are 14,000 babies that will die in the next 48 hours unless we can reach them.”

He was appealing for aid to be allowed into the Gaza Strip, after almost three months of a total blockade by Israeli forces which saw the Palestinian enclave come close to famine. The comments came during a week of growing international pressure on Israel to allow the free flow of aid. On May 19, Israel announced it would allow limited aid into Gaza, but the first day only saw five trucks cross into the Strip.

Nathan Gallo from France 24’s The Observers writes that when asked for clarification by the BBC, the UN’s OCHA said the figure came from an IPC report.

“In its latest report on the Gaza Strip, published May 12, the IPC states that the enclave is ‘still confronted with a critical risk of famine’, adding that ‘the entire population is facing high levels of acute food insecurity, with half a million people (one in five) facing starvation’,” Gallo writes.

“As to the number of malnourished children, the IPC reports that 14,100 children aged 6 months to 5 years old are actually experiencing ‘severe acute malnutrition’.”

So what about the 48 hours statistic? France 24 contacted the IPC, which clarified the statement, saying: “If they don't get access to enough food, these children are at risk of death if they're not treated over a one-year period.”

Armêl Balogog from Radio France’s Vrai ou Faux writes that “the IPC states that 14,100 cases of severe acute malnutrition are likely to occur among children aged 6 months to 5 years between April 2025 and March 2026, i.e., in one year, if the situation does not change. This is not a risk of imminent death, but it is still an extremely critical situation.”

Balogog writes that the UN office explained to the BBC that they needed to save an estimated 14,000 severely malnourished babies, and that they need the aid as quickly as possible, ideally in the next 48 hours. “During his interview, Tom Fletcher must have gotten confused by all these figures,” Balogog surmises.

“Since then, the UN's appeal has been heeded by Israel. Humanitarian aid was able to resume in Gaza on the night of Wednesday to Thursday, after two and a half months of total blockade, but it is trickling in. Time will tell if this will be enough to save these 14,000 young children from starvation.”

How Russians are shopping for Swedish luxury

Since Russia’s full-scale invasion of Ukraine over three years ago, it has become much harder for Russians to buy products from Europe.

The EU responded to the war with 17 different sanctions packages. Online shopping has been severely restricted, and Russian bank cards were excluded from the international SWIFT system — a measure aimed to stop money flowing to Russian companies and make trade with the West more difficult.

On top of all this, sanctions were placed on the export of luxury to Russia, covering goods like shoes, clothing and bags costing over €300 or electronics valued at over €750.

Despite all this, people inside Russia are still buying luxury goods, including from Sweden, Maria Lapenkova writes for SVT Verifierar. But just how are they doing it?

One method is the use of personal shoppers, who buy the goods abroad and ship them to Russia. Lapenkova was able to find thousands of these accounts when searching a popular hashtag for the service.

Another method is a service offered by Russian courier company CDEK, which helps people to buy goods located outside Russia and delivers them on request.

Lapenkova writes that the Russian government has also found ways to circumvent trade sanctions. In May 2022, Russia legalised parallel imports, which means that companies can import Western goods without approval from the manufacturers. This applies to a wide range of goods, as long as they enter via another country.

“This loophole is used by Russia's largest online stores, Wildberries and Ozon, which buy foreign goods in large quantities through parallel imports and stock them. Russians who shop in Russian online stores therefore buy products that are already available in Russia - making the shopping experience as smooth as any domestic e-commerce.”

Although parallel imports don’t technically violate EU sanctions, they could undermine the purpose of the restrictions, Lapenkova writes. For Western brands, it also means their products may still be ending up on Russian shelves without their will or even knowledge.

The close second place won by the Israeli artist led to widespread backlash about the voting system and fresh calls for the country to be expelled from the song contest. An investigation showed how the Israeli government leveraged a state-funded online advertising campaign for the country’s entry.

The issue did not fully bubble over until after the Grand Final was over. However, even earlier in the week, a simple incorrect message reply to voters in one of the semi-finals raised the heckles of fans, forcing the contest’s organisers to clarify that all votes were indeed counted correctly.

Meanwhile in the Middle East, construction sites appeared in Gaza that appeared to be part of an Israeli plan to deliver a controversial new aid plan. The Israeli government fully blockaded the Gaza Strip since March, leading to a mass hunger and health crisis in the enclave as attacks continued.

As activists, humanitarian organisations, and world leaders called for aid to be allowed into Gaza, some fake news circulated about aid drops over the strip, which didn’t actually take place.

Here’s our fact-check dispatch.

Israel in focus at Eurovision Song Contest

The 69th edition of the Eurovision Song Contest kicked off last week in Basel, Switzerland, ending with a Grand Final on Saturday night, May 17.

With Israel taking part for a second year since the launch of its devastating war on Gaza, the country’s act Yuval Raphael and the Israeli delegation were again under the spotlight.

The contest was ultimately won by Austria’s JJ, but the huge public vote for Israel’s Yuval Raphael put her into second place. Theories about the vote began circulating, with fans wondering how the act gained such huge public support from countries where the song seemed neither domestically popular nor support for Israel particularly high.

The story evolved in the days after the Grand Final, with several member broadcasters calling for a discussion with the EBU about both Israel’s participation in the contest and the televoting system. It even reached the highest political level when Spanish Prime Minister Pedro Sánchez made a public statement saying Israel should be removed from the Eurovision in line with Russia’s expulsion amid the war in Ukraine.

The organisers of the Eurovision Song Contest confirmed that there were no irregularities with the voting. There is also no rule against paid advertising or promotion for the artists, a tactic some countries and delegations lean into more than others.

An investigation by Eurovision News Spotlight uncovered evidence that an agency of the Israeli government deployed cross-platform advertising and used state social media accounts to directly target and encourage the public in participating countries to vote for the Israeli entry in the Eurovision Song Contest.

The Israeli Government Advertising Agency, which serves various government offices, corporations, government-owned companies, and publicly-owned enterprises, placed advertisements across Google products. They also provided instructions on how to vote up to 20 times for the country’s representative in the 2025 contest held in Basel, Switzerland.

The advertisements placed by the Israeli agency garnered millions of views across the internet leading up to Saturday’s final in Basel.

The inclusion of advertising and promoting the country’s Eurovision Song Contest entrant comes as a departure from the agency’s regular activities of communicating the Israeli government's perspective on the Gaza conflict, its defence policies, and its relationships with international bodies.

Analysis of the Google Ad Transparency Center by the Spotlight team found that the Israeli Government Advertising Agency has run more than 5,000 advertisements, some of which target the credibility of organisations that criticise Israel, counter accusations of wrongdoing, and canvass support for Israel's position on issues such as its military operations in Gaza.

A statement from Martin Green, Director of the ESC, said: “The Eurovision Song Contest's rules are designed to ensure a fair and neutral competition. These rules do not prohibit participating broadcasters or third parties such as record labels or others from promoting their entries online and elsewhere, as long as such promotion does not instrumentalize the Contest or breach its editorial guidelines. Many delegations employ paid promotion campaigns to support the song, profile, and future careers of their artists.”

The investigation was also covered by the team at Spain’s VerificaRTVE, Daan Nicolay of Belgium’s VRT NWS, and Switzerland’s SRF.

Voting message error sparks theories

Fans in Norway reported getting a strange message when they tried to vote for Croatia in the first semi-final. Translated into English, the message said: “Thank you for voting for Israel! You can vote 20 times, so if you like the entry, you can vote again. Best regards Eurovision 2025.”

Eurovision News Spotlight sought a statement from the organisers on what exactly happened here. Martin Green, Director of the Eurovision Song Contest, said: “Our voting partner Once has confirmed that a valid vote was recorded in all countries participating in this year’s First Semi Final. Some voters in Norway received an incorrect message when voting but votes for all countries were correctly attributed and this did not impact the result.”

Based on this, the automated message error was likely a technical mix-up between act 14 on Tuesday and act 14 on Thursday.

The incident was a reminder that while this glitch could have affected any of the participating countries, the fact that it involved Israel heightened suspicion, since it came amid many fans’ broader distrust in the organisers over their handling of participants whose countries are engaged in a conflict. It’s unlikely the error would have caused such a stir had it involved two more neutral participants.

As always, when analysing claims, it is important to look at the context and viewpoints of those proliferating them, as they may give us vital clues that point to confirmation bias.

Aid distribution centre construction sites appear in Gaza

Israel is preparing a number of sites in Gaza that could be used as distribution centres for aid under an Israeli-U.S. aid plan that was rejected by the United Nations, according to a report by BBC Verify, which analysed satellite imagery.

Benedict Garman, Matt Murphy and Merlyn Thomas wrote how the U.S. had earlier confirmed that it was preparing a new system of providing aid from a series of hubs inside Gaza, “which would be run by private companies and protected by security contractors and Israeli forces”.

UN agencies said they would no co-operate with the plan, saying it went against fundamental humanitarian principles. “A spokesperson for the UN's Office for the Co-ordination of Humanitarian Affairs (OCHA) accused Israel of seeking to use ‘food and fuel as leverage, as part of a military strategy’.

“All aid would be channelled through a handful of militarised hubs,” Olga Cherevko told BBC Verify. “That kind of arrangement would cut off vast areas of Gaza – particularly the most vulnerable, who can't move easily, or are otherwise marginalised – from any help at all.”

But do we know whether the plan will really be going ahead? Well, the team at BBC Verify analysed a series of recent satellite images from Planet Labs, where they were able to see new clearings and the beginnings of construction sites being established.

“BBC Verify used satellite imagery to identify four potential sites based on the limited available information about their locations. The sites are similar in size, shape and design to existing open-air distribution sites inside Gaza, such as at Erez, Erez West and Kisufim. The largest site we've looked at is bigger - more comparable to the area inside Gaza at Kerem Shalom crossing,” they wrote.

Analysis of a high-resolution image captured on May 8 showed bulldozers and excavators working on a section of land near the Egyptian border. Other images from May 11 and 12 show this and three other sites continuing to expand. “One site is about half a kilometre from a collection of eight UN warehouses, and 280m from another large warehouse.”

Three of the four sites identified by BBC Verify are south of the IDF's newly created Morag Corridor.

China didn’t drop aid into the Gaza Strip

The fact-checkers at Germany’s WDR looked into a story for their social media explainer channel, tickr news, about alleged aid air drops into Gaza amid the total blockade on the enclave.

The story began when they noticed videos circulating claiming to show Chinese aid drops over Gaza. The clips showed aid being dropped from aircraft by parachute, similarly to how many countries have delivered aid to the Palestinians since the beginning of the war.

Many of the comments were complimentary about China’s efforts to deliver aid. But, as the presenter Tom Schachtsieik explains, these videos are misrepresented.

“Here’s the kicker. We really do see a Chinese plane, which was also used to deliver aid like this. But in Afghanistan, for example, not Gaza.”

Another of the clips showed Chinese planes flying over the pyramids of Egypt, saying they were on their way to Gaza. “The truth is that the footage is from a military exercise conducted by China and Egypt at the beginning of May. No Chinese aeroplanes flew over Gaza,” the WDR team said in its analysis.

False aid delivery stories have the potential to hamper the world’s understanding of the real situation inside the Gaza Strip, where the UN says the entire population is now at “critical risk” of famine, further highlighting the importance of this fact-checking work.

Can Trump accept $400m gift?

In the US, much of the news cycle over the last week centred on Donald Trump and a gift from Qatar worth $400 million. The present? An ultra-luxury Boeing jumbo jet to be used as Air Force One.

Opposition politicians slammed the administration’s intent to accept the plane as “wildly illegal”, while even some of Trump’s own supporters were not in favour of the president accepting such a lavish gift.

For BBC Verify, Jake Horton, Tom Edgington and Joshua Cheetham looked into the actual laws around gifts that can be accepted by the president.

Experts in American history and constitutional law told BBC Verify that laws around gifts were brought in to prevent bribery or influence the government, and this latest proposed gift is “certainly stretching the Constitution”.

Horton, Edgington and Cheetham write: “There have been a number of other laws passed by Congress relating to the acceptance of foreign gifts, such as the Foreign Gifts and Decorations Act of 1966, which means that congressional consent is required for the acceptance of foreign gifts above a certain value.”

Generally, U.S. officials are only allowed to accept gifts valued under $480.

There was also a question around what might happen to the plane after Trump leaves office. “Although Trump has referred to the plane ultimately going to his ‘library’, experts have suggested he really means his museum foundation. Ex-presidents typically have a library housing their archive of documents, and a museum - typically funded by private donations - full of memorabilia and open to the public.”

“Experts who BBC Verify spoke to said the fact that the plane could be given to the administration - and not to the president directly - before being transferred to his museum, may not get around the potential violation of the constitution.”

It’s not to say that Trump may not still go ahead and accept this extravagant gift; the U.S. President has been known to push the limits of his powers. However, as we’ve seen in past dispatches, expert insights around complex and obscure issues are invaluable for both journalists trying to cover the story and the public at large.

And in such a divisive information environment, the Trump plane story generated dozens of counterclaims among the right-wing online community, which were debunked by fact-checking outlets. They saw users posting about a $17m helicopter Bill Clinton was allegedly gifted by Saudi Arabia, and a supposed gift by Kuwait of a $90m luxury jet to Al Gore.

India’s military forces said its May 7 operation, codenamed “Operation Sindoor”, was a retaliatory response to the Pahalgam attack, for which it claimed Pakistan was responsible, a charge the country strongly denies.

Pakistan described the subsequent offensive on its territories as an “unprovoked” attack. Pakistan’s Prime Minister Shehbaz Sharif said the "heinous act of aggression will not go unpunished".

The operation gave rise to a wave of content online, with many videos and images purporting to show military action by both sides. But could the content be trusted?

Meanwhile, senior Russian officials, backed by pro-Russian Telegram channels, promoted claims that French President Emmanuel Macron took cocaine with German Chancellor Friedrich Merz and British Prime Minister Keir Starmer while on a train travelling to Kyiv.

Wild conspiracy theorists claimed Macron was concealing a bag containing cocaine while Merz was holding a white cocaine spoon.

Is it true, or just another Russian disinformation campaign around drug-taking to marginalise European efforts to support Ukraine?

Here’s our fact-check dispatch.

Old videos and gameplay footage circulate amid India-Pakistan conflict

Old content goes viral

Amid the flurry of fresh conflict, fact-checkers and OSINT experts found lots of old content from other wars in circulation, wrongly captioned and misrepresented.

Derek Bowler from the EBU found an example of a Gaza war-related video in use, purporting to show the events in Pakistan and Kashmir. A video emerged that was captioned in several posts as showing the moment of strikes in Bahawalpur, a city in Punjab, Pakistan.

A reverse image search on Google Lens revealed that an exact match for the video had previously been published in October 2023, with a caption about intensifying Israeli attacks on Gaza.

Similar examples were found by other fact-checkers. At Deutsche Welle, Jan D. Walter and Sarah Steffen debunked videos that purported to show missiles raining down on Pakistan, but which actually showed Iranian missiles hitting Israel. The clips bore the logo of state-owned Indian news channel DD India, but it was proven to simply be their international coverage of the Middle East conflict, since the channel itself had published the video on October 1, 2024.

Another video probed by the Deutsche Welle team had been posted several weeks earlier with the hashtag ‘Gaza’. A further hint that it had no relation to the India-Pakistan conflict came via the audio, in which people could be heard speaking Arabic.

Matt Murphy, Olga Robinson and Shayan Sardarizadeh from BBC Verify found a video that had racked up 600,000 views on X and claimed to show that the “Pakistan army blew up the Indian Brigade headquarters”. In fact, they write: “The clip, which shows blasts in the darkness, is actually from an unrelated video circulating on YouTube as early as last month.”

The fact-checkers at VerificaRTVE did a deep dive on misinformation that sprouted up amid the fog of war, finding examples of old and misrepresented content that actually showed places in the U.S., China, Lebanon, and Gaza.

In one case, a claim that a clip showed a nuclear explosion in Pakistan was actually a video of a blast in Tianjin, China, in 2015, as confirmed by a reverse image search and news articles from the time.

Another example involved a video with claims it showed the floodgates of India’s Kishanganga Dam breaking after India tried to stop the flow of water into Pakistan. This couldn’t have been further from the truth; it was actually recorded in Texas in 2019, when the “Guadalupe-Blanco River Authority experienced a floodgate failure”.

And in the Middle East, it wasn’t just Gaza videos in the frame. A video showing an explosion that was posted with claims about a Pakistani counterattack on India was actually from Beirut in 2020, RTVE found after a reverse image search.

The footage was filmed during the port explosion in the Lebanese capital on August 4, 2020, and was used extensively by reputable news outlets to illustrate the story way back then.

Gameplay videos presented as real

This fresh conflict also generated an often-seen phenomenon of gameplay clips misrepresented as real war footage.

In one such case, a video purported to show surface-to-air missiles and guns downing Indian and Pakistani jets. The EBU’s Bowler noted how the videos were all similar, with a purple-blue skyline and weaponry trails appearing highly saturated.

A reverse image search brought him to the profile of a gaming streamer, who regularly posts game content depicting war scenes. The same video was posted in higher quality on the gaming page in April. It was confirmed by that page as being Arma 3 gameplay. Arma 3 is a tactical shooter simulation game, and clips from it frequently come up in fact-checks about real conflicts, such as those in Ukraine and Gaza, due to its highly realistic graphics.

In another case, the team at BBC Verify found a set of photos purporting to show an operation by Pakistan’s Air Force targeting Indian air bases on May 6. Those pictures were actually taken from the video game Battlefield 3.

Deutsche Welle also found video game footage in circulation. They carried out a visual analysis, where they saw that the video was suspiciously well-lit and had an artificial feel to it. “Going through the video frame by frame, the unnaturally pixelated lights are a hint that the video is artificial,” Walter and Steffen write.

Old plane crash content reshared

BBC Verify looked into a series of clips claiming to show the wreckage of Indian fighter jets, after the Pakistani military said it destroyed five aircraft in the fracas. Some of them had garnered millions of views.

Two images that were widely shared showed previous crashes involving Indian Air Force jets: one in Rajasthan in 2024 and another in Punjab in 2021.

At Deutsche Welle, Walter and Steffen also looked into an image from the 2024 Rajasthan crash, as well as finding an example of a presumably pro-Pakistan user who posted an image of what they claimed was a French-made Rafale fighter aircraft, in use by the Indian Air Force, which was shot down during a strike near Bahawalpur.

The image did indeed show a French-designed aircraft, but it was actually a Mirage 5. “This type of assault jet was retrofitted for the Pakistan Air Force so that it could carry and deliver nuclear weapons,” the DW team writes. According to local media, the wreck seen in the image was actually a plane that was on a training mission in the Pakistani province of Punjab three weeks earlier, when it crashed. Both of the pilots survived with minor injuries after ejecting themselves.

Pro-Kremlin accounts spread conspiracy theories with false claims of drug use by Macron, Merz and Starmer

Leaders from four European allies visited Kyiv on May 10 in a show of support as they called for Russia to agree to a 30-day ceasefire of its war in Ukraine.

French President Emmanuel Macron, German Chancellor Friedrich Merz, British Prime Minister Keir Starmer and Polish Premier Donald Tusk arrived by train in the Ukrainian capital on Saturday, where they met Ukrainian President Volodymyr Zelenskyy.

However, it was the train journey itself that would become a hot topic online, with pro-Russian Telegram channels claiming a video released of Macron, Starmer, and Merz onboard the train travelling to Kyiv showed evidence that the three were taking cocaine.

A video circulating on social media over the weekend of the trio's meeting aboard the train shows Chancellor Merz pushing away a glass of water and a blue folder, then grabbing a sharp object in his right hand. President Macron is seen picking up a white object and holding it while posing for photographers, before discreetly putting it in his pocket.

The pro-Kremlin narrative

Pro-Russian accounts suggested that President Macron was concealing a bag of cocaine on a table aboard a train, while others speculated that Chancellor Merz was holding a cocaine spoon.

On Sunday, the Kremlin's foreign ministry spokesperson, Maria Zakharova, amplified the claim on her Telegram channel, writing: “After pushing Zelensky into yet another hellish scheme to derail a settlement and prolong the bloodshed in Europe, it's like a joke: a Frenchman, an Englishman, and a German got on a train – and did a line.”

Zakharova shared low-resolution images from a video of the three, which included red circles of objects on the table, suggesting it proved the theory of drug use by the respective leaders.

“Apparently, they were so out of it they forgot to hide the paraphernalia – a small bag and a spoon – before the journalists arrived. Europe's fate is in the hands of placeholders who are dependent, in every sense of the word,” Zakharova wrote.

The head of the Russian Direct Investment Fund (RDIF) and an envoy of Vladimir Putin, Kirill Dmitriev, added to the speculation, commenting: “Is this footage AI or real? If it's real – are we looking at sugar or something entirely different? If it's something else, it explains a lot of recent ideas and proposals.”

French conspiracy theories

The theories of drug use aboard the train quickly spread to other nations, including France, where Linh-Lan Dao of France Télévisions investigated commentary on the video by two far-right figures.

“Florian Philippot, Marine Le Pen's former right-hand man, now head of the Les Patriotes party, wrote on X: ‘When the photographers arrive in the middle of their sleepover, the three war-mongering idiots seem uncomfortable: what is Macron hiding?!’”

The official X account of Debout La France, a French political party founded by former presidential candidate Nicolas Dupont-Aignan, shared a post containing the video, writing: “But what is Emmanuel Macron hiding with the air of a teenager caught in the act?”

“The two far-right figures, who are politically close, are known for relaying disinformation and pro-Kremlin narratives,” Dao writes.

“The @Veritiste account, who claims to be a ‘conspiracy theorist’, even mentions a ‘cocaine evening among friends’, believing that the object seized by the German chancellor was a ‘cocaine spoon’,” Dao adds.

A handkerchief

Nele Baeyens and Daan Nicolay of VRT NWS compared the low-resolution images to those filmed and shot by the Associated Press.

“In the images that are going viral, it is not entirely clear what the object is. However, other images of the meeting remove all doubt. In photos and videos from the AP news agency, you can clearly see that it is a crumpled handkerchief. The spoon that the German chancellor is said to have quickly hidden looks more like a crumpled cocktail stick in those images,” the pair write.

Paula Peña of VerificaRTVE also examined similar photographs shared by AFP and Getty Images, writing: “Higher-resolution images from news agencies support the Élysée Palace's clarification. France-Presse (AFP) and Getty Images ( 1 and 2 ) have published several photographs of the train meeting between Macron, Starmer, and Merz, in which it can be seen that the white object in front of the French president is a white napkin.”

In response to the claims, the Élysée Palace said: “This fake news is being spread by France's enemies abroad and at home. We must remain vigilant against manipulation.”

“This is a tissue. For blowing your nose," the Élysée posted next to an image from the meeting. “This is European unity. To build peace. When European unity becomes inconvenient, disinformation goes so far as to make a simple tissue look like drugs.”

A history of cocaine-related disinformation

Austrian public service broadcaster Österreichischer Rundfunk, in its reporting of the train video, said that the claim about the cocaine-snorting head of state fits with other pro-Russian narratives.

zeitimbild

A post shared by @zeitimbild

ORF quoted an expert from the Center for Monitoring, Analysis and Strategy (CeMAS) as saying: “The Western heads of state are thereby meant to be discredited and portrayed as untrustworthy. Narratives surrounding drug use have been used by Russian propaganda for years to discredit Ukraine and especially President Zelenskyy.”

In April of 2022, Catalina Marchant de Abreu of France 24 debunked digitally altered videos on social media from pro-Russian accounts, which claimed that President Volodymyr Zelenskyy was addicted to cocaine, concluding the rumours dated back to 2019, during Zelenskyy's presidential campaign.

Social media users accused Zelenskyy of being “coked out of his mind”, “antsy”, and “high on more than life” during an interview with NBC News in February 2024.

Deputy Chairman of the Security Council of Russia, Dmitry Medvedev, in December 2024, accused the Ukrainian president of being a drug addict who would prolong the Russo-Ukrainian War to maintain his hold on power.

“The regime will continue to insist on a total war to the last Ukrainian and drag out the elections as much as possible...Because this is the only way to preserve the power of an illegitimate drug addict,” Medvedev shared on Telegram at the time.

Key takeaways

• Reverse image search and source tracing remain powerful tools for debunking reused or mislabelled content.

• Low-quality media is often used strategically to seed disinformation — always seek alternative angles or clearer versions.

• Verify language, location indicators, and publication timestamps to contextualise content.

• Misinformation often follows predictable patterns (e.g., drug-use tropes and recycled footage) — recognising these patterns can speed up debunking.

• Collaboration between fact-checkers, journalists, and open-source analysts is crucial in restoring factual clarity.

The blackout, which lasted for much of the day, brought daily life to a standstill, with transportation systems, hospitals, and communication networks across the Iberian Peninsula among the most severely affected.

Spain’s grid operator, Red Eléctrica de España (REE), initially ruled out a cyberattack but pointed to a failure stemming from two solar plants in the southwest of the country, which destabilised the power system and led to a cascading breakdown in the interconnectivity with France.

While electricity has been restored, the cause of the outage remains under investigation. As authorities combed for answers, the incident gave rise to a wide range of misinformation online, from misrepresented content to wild conspiracy theories emerging about the cause of the outage.

Here’s our fact-check dispatch.

Altered space images used in coverage of blackout

As social media users and even news organisations scrambled to find images to illustrate the power outage, several photos captured from space showing the Iberian Peninsula in darkness came to the fore.

Whether it’s smoke from burning wildfires, urban lighting visible from space, or damage after a natural disaster, satellite imagery and photographs shared by astronauts can be an invaluable asset. But, of course, only when they’re correct.

One of the images shared soon after the power cut appeared to have been shot from the International Space Station. It showed Spain in darkness, with lights in other countries clearly visible.

The image was even run by an Argentinian news outlet as part of its coverage of the story.

Derek Bowler from the EBU analysed the image and found a few interesting things. The lights in Spain were visible, but darkened compared to neighbouring countries. The picture also showed a perfect cutline on the border of Spain and Portugal, despite the lights being visible on both sides. This indicated that a darkening filter had been applied to Spain.

Using Google Lens, Bowler was able to find an article by NASA from 2014, showing the original image in which Spain was illuminated just like anywhere else. It was captured in July, 2014, by crew members on board the International Space Station.

Other images were also used similarly. VerificaRTVE in Spain, which checked a lot of the content being widely shared around this domestic story, found this image apparently captured from space that was falsely captioned as showing the blackouts.

The RTVE team found that the image was likely AI-generated, through analysis via two AI detection tools. The image also shows France in complete darkness, which did not happen during the power cuts that mainly affected Spain, Portugal, and parts of southern France.

And there was yet another satellite image being shared around, this time showing all of Spain and Portugal in darkness. It also showed the Balearic Islands without power, even though those islands were unaffected by the blackout.

The team at RTVE picked up on another detail that proved the image was not genuine. Night fell at 9:32 p.m. on the night of the power cut, and by then, about 35% of the power supply in A Coruña had been recovered. “If this were a real picture, lights should be visible in that area,” the team at RTVE write.

VerificaRTVE also debunked several other instances of misinformation that emerged during the power outages, from fraudulent scams to circulating messages that falsely claimed Ursula von der Leyen had called it a Russian attack. There were also suggestions that Israel or the U.S. were behind the blackout, but no evidence was provided.

The proliferation of these types of images show just how easy it is to fake or alter a space image to depict a power outage, and why we should be on the lookout for this genre the next time a similar story emerges.

Did too much solar power trigger the blackout?

If you were following news of the power outage, you may have come across screenshots of graphs and diagrams showing fluctuating lines and a nosedive just as the electricity went out.

Many of these were shared with claims about Spain’s reliance on renewable energy, specifically that an oversupply of solar power may have triggered the sudden failure.

The claim was posted by someone described as a physicist, who wrote: “Renewables, in this case solar, have just caused the first major #power outage in Spain, specifically due to an oversupply of solar.” That post included screenshots which cited German chemistry professor and energy transition critic, Fritz Vahrenholt, who said that renewable energy production in Spain had surpassed demand before the grid outage.

Jan D. Walter looked into the theory for Deutsche Welle. He writes that the head of the Spanish grid operator REE, Eduardo Prieto, addressed other claims in ruling out a cyberattack in the immediate aftermath of the blackout, but gave some further analysis on what actually happened leading up to the incident.

“According to Prieto, two separate incidents in southwest Spain within 1.5 seconds apparently caused the grid frequency to become so unbalanced that power plants with a capacity of 15 gigawatts (GW) failed in a very short time - almost half of the power plant capacity active at the time.”

Walter goes on to write: “Very large and sudden deviations can lead to a cascade of errors. An extreme frequency fluctuation triggers mechanisms at other power plants (and large consumers), which then automatically disconnect these plants from the grid.” Prieto said it was “very possible” that this fluctuation could have been triggered by solar power plants.

However, as Walter writes, the cause of the power outage remained unclear and it was not yet possible to prove that an oversupply of solar energy caused the blackout. It is true that solar energy was providing some 60% of the electricity in the Spanish grid, and solar plants seem to dominate electricity production particularly in southern Spain, where the faults were recorded.

Since then, new information has come to light. On Monday, Environmental Transition Minister Sara Aagesen told Spanish national broadcaster TVE that there was a “third loss of power generation” that came 19 seconds before the blackout, adding to the two that were already known about. The grid withstood this first fluctuation, but not the other two that came five seconds before the outage.

The minister said: “So far, we know that this new incident 19 seconds earlier occurred in southern Spain, and the following two occurred in the southwest.” The government is not ruling out a cyberattack.

It’s difficult for journalists to delve into such a scientific topic when all of the information is not yet clear. But as this story shows, what we can do is use the information that is publicly available, to at least show that the claim can’t yet be proven.

Miscaptioned content amid power cut panic

As with most major breaking stories, the power outage also generated lots of social media posts featuring old, fake, and miscaptioned images.

For Belgium’s VRT nws check, Nele Baeyens, Ellen Debackere and Daan Nicolay looked into some of the content that spread around platforms in the hours after the incident.

One such photo showed people resting on an escalator, with a caption: “Image of Belgians stuck on an escalator due to the power outage in Europe.” The post had over 200,000 views. But it wasn’t true. A reverse image search showed it was already in use in France in 2021.

Other videos showed black smoke spewing out of chimneys, purporting to show explosions at a refinery in Biscay, and accompanied by claims about a “massive cyberattack”.

VerificaRTVE researched the incident, finding that Petronor, the company operating the refinery, published a statement clarifying that they were standard “safety measures” to release pressure.

There was more detail from the local fire department, which confirmed it recorded the smoke emissions and said they were “not accidents”. The Bomberos Bizkaia released this footage showing the situation, writing: “With the widespread power outage there have been many #rescues of people trapped in lifts and calls of industrial accidents. Companies put their production processes in ‘protection’. They are not accidents, it’s safety.”

The RTVE team further cited local reporting that said the refinery units were shut down due to the power outage, and then the safety measures were activated.

But that wasn’t all. A photograph also emerged of Spanish Prime Minister Pedro Sánchez and ministers gathered around a meeting table, claiming it showed a committee meeting over the power cut.

It actually shows a Council of Ministers meeting in January, and has no connection to the blackout. The reporters at RTVE were able to verify this by conducting a reverse image search, where they found the image published by the Spanish National Council. There was a meeting convened to deal with the power cut, but this wasn’t it.

The use of unrelated images and the incorrect captioning of real footage is nothing new, and comes up whenever we see a major story break into public discourse. As these stories show, sometimes all you need is a critical eye, some healthy suspicion of everything online, and a reverse image search.

The masses of material ranged from AI-generated homages to the popular pope, to fake biopic trailers, to all-out conspiracy theories.

While we would always expect stories of this magnitude to spawn fakes and false claims, the situation was heightened given the perceived secrecy, intrigue and already-existing conspiracy theories that have long surrounded the Vatican and its inner workings.

On top of this, Francis was a leader that in some circles was considered even more controversial than other popes. He was seen as an outspoken progressive figure who was frequently criticised publicly by conservatives both inside and outside the church.

That’s without even mentioning the sectarian divides that have hounded the Christian world stretching back hundreds of years.

All this led to the perfect storm for misinformation in the wake of Pope Francis’ death.

Secret satanic signs and apocalyptic predictions

A massive web of linked conspiracy theories gained prominence in the days after the pope’s death, some involving purported secret signs that the Catholic Church is really satanic, and others about apocalyptic predictions and prophecies involving old societies and orders.

You’d be forgiven for thinking it all sounds utterly bizarre. But for die-hard conspiracy theorists, these beliefs run as deep as any religious creed.

One of the claims involved a statue that is a main feature in the papal audience hall in the Vatican. ‘The Resurrection’, a colossal work in bronze made by Pericle Fazzini, is regularly seen in coverage of the pope as it is right behind his seat in the Paul VI Audience Hall, indoor events for large groups are typically held.

Claims that were already out there before the pope died were themselves resurrected in the past week, saying that the statue featured Jesus surrounded by demons and with a snake coming out of his head.

But that’s not true. The artist himself explained the 1970s piece in various writings about his work. It was partially inspired by the nuclear bombing of Hiroshima, and more widely the looming threat in the 20th Century of nuclear war. The huge statue shows Jesus rising from a bomb crater in the olive groves of the Garden of Gethsemane, surrounded by destruction.

Other theories swirled about the audience hall building itself and St Peter’s Basilica and the adjoining square were designed to look like snakes, the symbol of evil in Christianity going back to the earliest story of Adam and Eve’s temptation by a serpent.

There was no basis for the claims, and those sharing them used images captured with a fish-eye lens to support their argument.

The team at VerificaRTVE probed a number of cases of misinformation, some of which also revolved around the satanic theme. A video posted on X and Instagram showing a screen recording of the arrival of the pope’s coffin at St Peter’s Basilica was accompanied by captions claiming it proved a satanic link.

The RTVE team write: “We then see the doors leading to the steps next to the main altar, which descend to St. Peter's tomb, and the shot closes in on two inverted crosses.”

VerificaRTVE verified the footage, which matched that of the official broadcast of the event. As for the claim, they write: “The inverted crosses that appear in the video are not a satanic symbol, but the Cross of St Peter … According to traditional iconography, Saint Peter was crucified upside down on an inverted cross, at his explicit request, as a demonstration of his inferiority to Christ.”

The upside-down cross actually appears in various Vatican documents, the RTVE fact-checkers write. They point out that they have in recent months also debunked other theories alleging that Pope Francis and his predecessors had propagated “satanic symbols”.

Rayna Breuer, Aldo Sanchez Vera, and Kathrin Wesolowski from Deutsche Welle also investigated a wide range of false and unproven claims that took aim at the pope’s reign, including posts about him congratulating Putin on his election win, calls to euthanise disabled people to curb climate change, and his alleged collaboration with Argentina’s junta in the 70s and 80s.

It would be impossible to unpick the huge web of conspiracy theories that have been circulating for decades about the Vatican and whoever sits at its head. One theory often leads into another, leaving proponents in a rabbit hole that can quickly become entirely detached from reality. Analysing the visual material accompanying these claims and consulting experts can be an important way for fact-checkers to at least challenge some of these baseless suppositions.

AI imagines pope reaching heaven with other deceased celebrities

AI video enthusiasts were busy during the week imagining how it might look to see Pope Francis in heaven. There were also fake movie trailers, including one for a ‘Disney’ production that saw the Argentine pontiff being played by a Black woman.

In one video (archived), he is approached and embraced by Britain’s Queen Elizabeth II, with Michael Jackson, Amy Winehouse and Silvio Berlusconi also in the frame. In another clip (archived), the pope flies through the skies with Jesus Christ, takes selfies with him and Mary, and marvels as he is suddenly able to turn water into wine. It ends with an age-rewind that shows Pope Francis as a younger man. The Bob Dylan song ‘Knockin’ on Heaven’s Door’ plays in the background.

Another similar video (archived) seen by tens of millions of people shows Pope Francis in the clouds greeting other beloved celebrities who are dead, such as Princess Diana, Stephen Hawking, Pelé, Mac Miller, Liam Payne, and Robin Williams.

Another aspect of the AI-generated flood of material that emerged after Pope Francis’ death involved the technology’s use to mock up film trailers.

Footage was shared by multiple accounts showing a purported trailer for an upcoming Disney biopic about Pope Francis. In the viral clip, a Black woman plays the pope, and is seen facing crowds in St Peter’s Square and inside a chapel.

It might seem hard to believe that anyone would truly fall for these AI-generated videos, but as commenters pointed out beneath some of them (“my mom is going to think this is real footage”) there exists a large gulf in the public levels of media literacy, which can be particularly pronounced across generational divides. Secondly, keeping an eye on the types of AI-generated content circulating allows experts to monitor the ever-improving quality of these clips, which helps inform our approach in this area.

Pope Francis long-targeted by AI fakes

Pope Francis became famous for his own use of technology, becoming the first pontiff to have an Instagram account and frequently distributing messages to faithful followers via his Pontifex X/Twitter account.

But he frequently fell foul to AI-generated content purporting to show him, both before and after his death.

In one of the first most famous examples of an AI image that tricked the world, he was seen wearing a white puffer jacket instead of his usual cassock in 2023. Although the image itself was relatively harmless and made no wild claims about the pope’s reign over the church, it marked a watershed moment for AI, serving as a warning to the world about what this technology would go on to be capable of.

Despite his protests over the years, the fake news only ramped up when the pontiff became ill in February, prompting speculation about his closeness to death. Derek Bowler from the EBU documented the wave of AI-generated content that purported to show the pope in his hospital bed. No real pictures from the hospital bed were released, and these AI-generated fakes filled the void amid a massive public appetite for the story. Some of the material was found to have been created by Grok, a generative artificial intelligence chatbot developed by xAI, Elon Musk's AI company.

Armêl Balogog writes for franceinfo about instances in which the pope was critical of divisive disinformation and the era of ‘fake news’.

In January of this year, he told a gathering of journalists that too often today, communication “even uses false or artfully distorted information to send messages designed to agitate, provoke or hurt”. Balogog writes how Francis, years earlier, described fake news as “snake-tactics” and said that in the story of Adam and Eve, the “crafty serpent” was the author of the first fake news.