The Week of November 16, 2026
NewsGuard’s Reality Check. Dougan Leak Exposes Concerted Effort to Infect AI
Leaked documents show that John Mark Dougan, a U.S. fugitive turned Russian mouthpiece, tried to skew AI chatbot responses in Russia’s favor and regularly used chatbots to help write anti-Western ‘news’ reports. The leak also describes a network of 286 news sites that used AI to produce 6.3 million articles last year repeating 115 Kremlin false claims, about 17,000 stories a day designed to infect AI models.
The Canadian Press (via CP24). A deepfake video network is spreading false information about the U.S. trade war
A McGill University report finds that a network of nearly 300 TikTok and Instagram accounts, almost certainly run from outside North America, uses deepfakes of real influencers to spread politically divisive content about the U.S.-Canada trade war. The network has posted more than 3,201 videos seen 22.1 million times in three months, including one about a non-existent Trump ultimatum that drew more than five million views.
The Quint. No, Ex-Army Officer Pravin Sawhney Didn’t Call Pahalgam, Pulwama False Flag Ops
An AI-manipulated video circulating in India claims to show former Army officer Pravin Sawhney, in an interview with a journalist from The Wire, calling the 2019 Pulwama attack and the retaliatory Balakot airstrike ‘false flag’ operations. The Quint found the video was manipulated with AI and the man in it was misidentified as Sawhney.
Yahoo News. Fact Check: Viral Video Of People Sleeping In US Airport During Ground Stop Is NOT Real -- AI Made It
A Facebook video posted Nov. 14 claimed to show stranded travelers at a U.S. airport during the FAA’s nationwide ground stop of Nov. 13, but Lead Stories found it was likely made with AI. The clip showed two identical women in identical pink hoodies sleeping side by side, and passengers’ bodies appeared to merge into each other.
EUobserver. More EU sanctions on Russian propagandists and hackers, as thousands of fake accounts target voters
The European Union imposed new sanctions on 30 individuals and two organizations, including Moscow's Social Design Agency, for running cyberattacks and disinformation campaigns across Europe. Simultaneously, researchers from Alliance4Europe revealed that thousands of recycled Russian fake accounts targeted European voters ahead of recent elections in Sweden, the Netherlands, and Hungary.
CP24. AI deepfakes use Canadian influencers’ faces to spread political disinformation
A McGill University study uncovered a network of at least 293 TikTok and Instagram accounts stealing the likenesses of Canadian influencers and overlaying AI-generated voices to spread divisive political content. Over a three-month span, roughly 3,000 deepfake clips amassed more than 20 million views targeting the Alberta referendum, provincial votes, and the U.S. midterms.
AFP Fact Check. Misinformation about Colorado voter rolls spreads after 2026 midterms
Viral social media graphics falsely claim that Colorado's total voter registrations outnumber its adult citizen population as evidence of fraud in the 2026 midterm elections. Fact-checkers confirmed the misleading charts improperly include hundreds of thousands of inactive voters who have moved or died, while active registrations remain well below the voting-age population.
Yahoo News. Fact Check: No evidence Trump used 1950s immigration act to ban muslims from public office
Social media posts have spread a false claim that President Donald Trump plans to revive the 1952 McCarran-Walter Act to bar Muslims from holding public office. Fact-checkers found no evidence supporting the rumor and noted that the Cold War-era immigration statute could not legally or constitutionally be used to ban citizens from office.
University of Florida News. Deepfake faces appear “hyperreal.” These brain scans may explain why
Researchers at the University of Florida discovered a phenomenon called "deepfake hyperrealism," in which AI-generated faces are frequently judged as more realistic than photographs of actual humans. Brain scans linked this vulnerability to reduced activity in brain regions responsible for caution, particularly among older adults who are more susceptible to fake online identity scams.
Pursuit (University of Melbourne). Five ways to spot AI misinformation before you vote
Ahead of Victoria's November 28 state election in Australia, political parties are flooding social media and television with AI-generated deepfakes and cloned voices, a trend researchers call "misAInformation." Scholars warn that inconsistent platform labeling rules make it increasingly difficult for voters to distinguish genuine campaign messaging from synthetic political "ensloppification."
Deutsche Wells. Disinformation, misinformation and fake news are often intentionally spread for political purposes via social media platforms.
An investigation by DW Fact Check revealed that social media accounts posing as official representatives of the Iranian military have reached millions of users with coordinated propaganda. Analysts found that these accounts exhibit classic content-farm traits designed to amplify misleading wartime narratives across global platforms.
NewsGuard. 2026 Elections False Claims Tracker
NewsGuard's post-midterm update documented 70 viral false claims about the 2026 U.S. elections across 800 confirmed appearances on 12 platforms and 85 websites. Nearly half of the falsehoods focused on fabricated election fraud—including AI-altered candidate photos and bogus whistleblower accusations—with 68 percent of tracked instances appearing on Elon Musk's X.
NPR. Swatters are using AI voices to deceive 911 dispatchers, and it's only getting worse
Emergency dispatchers across the U.S. warn that bad actors are increasingly using AI-generated voices and synthetic background audio to make dangerous "swatting" hoax calls sound convincing. Law enforcement associations note that realistic synthetic audio makes it far more difficult for 911 operators to detect fabricated emergencies before dispatching armed officers.