8 Tactics Foreign Actors Used To Influence the 2026 Midterms

Influence operations linked to Russia, China, and Iran targeted America’s 2026 midterm elections. While goals differed, tactics overlapped: fabricated news reports, hijacked branding of trusted outlets and celebrities, and invented or purchased personas. Generative AI made the fakes easier to produce, and in at least one case, AI chatbots repeated them. Iran’s government also appealed directly to American voters. Lawmakers briefed on these operations saw no sign of foreign interference with the mechanics of voting itself. But midterms are often trial runs for new tactics. After struggling to get traction in 2024, these adversaries adapted, and what worked in 2026 will likely return in 2028. Here are eight tactics used this cycle: 1. Russia-linked operators fabricated abuse allegations against Democratic Senate candidates. A Russian influence operation known as Storm-1516 pushed fabricated news videos targeting three Democratic Senate candidates in the final weeks of the 2026 midterms. This tactic was designed to affect voter turnout by using polarizing issues to inflame anger or fear, rather than to convince voters of any single claim. The first false allegation used AI deepfake videos of a man claiming to be a former student who accused James Talarico of sexual assault. The real former student said the videos were “very much AI.” The second allegation used a video mimicking a Cleveland 19 broadcast and accused Ohio’s Sherrod Brown of sexual assault. The third impersonated Maine’s WABI in a segment accusing Troy Jackson of threatening staffers with an ax. Both outlets denied the reports: Cleveland 19 stated it reported no such thing, and WABI said its segment was fabricated. Collectively, clips of all three efforts reached 2.8 million views in the days after the allegations appeared. The fakes also exploited real Americans’ identities, including a former Brown intern. 2. A Russia-aligned network hijacked the credibility of trusted newsrooms and celebrities. A Russia-aligned network known as Operation Overload, Matryoshka, or Storm-1679 produced a set of videos in September that impersonated or falsely cited the BBC, CNN, Politico, and France’s VIGINUM (a government agency that counters foreign digital interference) to accuse Democratic candidates of corruption, antisemitism, and misconduct. Hijacking trusted media brands is a common Russian disinformation tactic. This approach depends on borrowed authority, using familiar logos and faces to lend credibility. The false reports can also damage the reputation of outlets and public figures they mimic. Overload also used authentic footage of actors, including Sarah Jessica Parker and Julia Roberts, with overlaid text falsely claiming they backed the “AllDemocratsAreCriminals” digital campaign. A separate operation posted nine fabricated news videos on September 17, claiming 10 celebrities had denounced Democrats over gender transition. Bloomberg reported AI-fabricated clips of Robert De Niro and Cher attacking Democrats, boosted by bot networks. 3. China- and Iran-linked networks invented American personas. Some foreign operations hide behind invented identities. The Chinese network Spamouflage posed as U.S. voters ahead of the 2026 midterms and posted divisive content about politics and the elections. Its personas drew little authentic engagement, and the aim was less to change votes than to undermine confidence in U.S. elections. Iran has used similar tactics. In late August, Meta removed an Iran-linked network of Facebook and Instagram accounts that pretended to be based in the United States and ran political meme accounts pushing anti-Republican messages, along with content about the Israeli-Palestinian conflict and immigration. The network also pitched content collaborations to political figures and news organizations, none of which succeeded. Still, roughly 79,400 Instagram accounts followed at least one of its meme accounts. 4. China- and Russia-linked networks exploited accounts that already had audiences. Building an audience for fake personas takes time, so some operators acquired accounts that already had one. China’s Spamouflage amplified a Philippine culture account with about 200,000 followers that had been purchased, hijacked, or otherwise repurposed — it is unclear by whom — to post almost entirely about U.S. social and political issues, including an AI-generated video arguing that America is an oligarchy. Followers did not guarantee engagement: the Philippine account’s posts typically drew only a few dozen likes and replies. Russia’s Doppelgänger influence network took a similar shortcut, likely supplementing its bot network with purchased or previously owned accounts registered in India, Singapore, and Hong Kong. Unlike its bare-bones bots, these accounts had custom photos, bios, and varied handles that made them look more like real users. 5. Iran- and Russia-linked operators posed as journalists and allied voices. In October, OpenAI revealed an Iran-based operation that used seven fake journalist personas, some claiming to be based in the United States and other Western countries, to place nearly 100 articles in roughly a dozen small and mid-sized online news sites. Its goal was to inflame public opinion and undermine support for the U.S.-Iran war. Russia’s Doppelgänger used a different disguise, operating more than 50 fake accounts posing as non-Americans, including Israeli and French voters. The Israeli personas pushed claims that Democrats had turned away from Israel. More than 1,000 bot accounts then quote-posted that content an estimated 150,000 times. Foreign operators hiding their hand is a hallmark of malign influence. What stands out this cycle is how they hid it: behind real news outlets and in debates Americans were already having. 6. Iran’s government appealed to American voters in the open. Iran did not hide all its efforts to influence American voters. On September 29, Islamic Revolutionary Guard Corps spokesman Gen. Hossein Mohebbi presented a 25-page “letter to the American people” at a news conference in Tehran. The letter urged Americans to oppose U.S. support for Israel and foreign wars and argued that the two countries could coexist if U.S. policy changed. Mohebbi urged Americans to take to the streets to protest President Donald Trump’s policies and said they should “revise their decision” about their leaders. He told reporters that he did not expect the letter to change the administration’s behavior but that voters might hear the message. 7. AI helped operators scale up, and it began running the operations. Generative AI enabled many of this cycle’s operations: Storm-1516’s Talarico deepfakes, Overload’s fabricated celebrity clips, the oligarchy video, Iran’s memes, and Doppelgänger’s AI-generated images. A few operators can now do the work of whole teams of writers, translators, designers, and managers faster and in more languages. With AI, operations can churn out more content that looks authentic — at lower cost. Meta says virtually every influence network it disrupts now uses some form of generative AI. A fabricated broadcast or persona no longer requires a large organization behind it, which lowers the barriers to entry and lets established operators run more campaigns at once. In 2024, AI mostly helped adversaries scale content. This cycle, operators began using it to run the operations. Foreign actors are deploying AI agents that create fake accounts, generate political content, and spread it with limited human oversight. Other operations have used AI models to build and run networks of fabricated news sites and social accounts. 8. AI chatbots repeated the fakes. AI changes how false claims reach voters, not only how the claims are made. The morning after the Talarico videos first appeared online, Google’s AI Overview, Gemini, and Microsoft’s Copilot all repeated the false claim. AI Overview cited ABC News and KTAL articles that never mentioned the allegation, so the fabrication acquired the appearance of sourcing it never had. Google indexed Doppelgänger’s fabricated articles, where chatbots could retrieve them in answers about the election. What Comes Next Foreign influence operations will not end with the midterms. Networks like Doppelgänger have reusable infrastructure that keeps their narratives circulating, and as AI tools improve, these campaigns will become harder to spot. The burden now falls on authoritative sources to quickly identify and attribute these operations and on voters to treat chatbot answers and potential deepfakes with appropriate skepticism. Making matters worse, the U.S. government has dismantled much of its capacity to attribute these operations publicly. In 2024, U.S. intelligence and law enforcement agencies quickly named Russia and Iran in multiple influence operations. This cycle, that job fell to researchers and companies. Private firms can trace a network, but only the U.S. government can tell Americans, with its intelligence and authority, which foreign government is behind it. Before 2028, Washington will have to decide whether it wants that job back. Jacob Breach is a senior advisor at the Foundation for Defense of Democracies (FDD), where he focuses on foreign influence operations. Pavneet Kaur is an intern at FDD’s Center on Cyber and Technology Innovation (CCTI). For more analysis from the authors and FDD, please subscribe HERE. Follow FDD on X @FDD and @FDD_CCTI. Follow Jacob on X @JBreach. FDD is a Washington, DC-based, nonpartisan research institute focusing on foreign policy and national security.