Donald Trump is facing a very different kind of election challenge as the 2026 midterms approach. According to US intelligence officials, Russia has launched a new online influence campaign designed to exploit America’s political divisions and undermine confidence in the electoral process.

The Kremlin’s latest effort is particularly striking because artificial intelligence is playing an increasingly important role, allowing Russian-linked networks to produce convincing-looking videos and fabricated political content at a speed and scale that would have been far more difficult during previous election cycles. U.S. officials say they have not found evidence that Russia is attempting to compromise voting machinery itself, but the online campaign is aimed at influencing what Americans see and believe during the election season.

And researchers watching the activity are warning that what has already appeared online may be only the beginning.

Researchers in Europe and the United States detected two separate waves of activity over the past week, with videos appearing on X and Bluesky and falsely or misleadingly targeting Democratic candidates in competitive Senate races. Some of the videos were reportedly designed to look as though they came from major news organizations, including CNN, the BBC and The New York Times, giving fabricated claims the appearance of legitimate journalism.

Graphika senior researcher Jean Le Roux offered a particularly blunt warning about where the campaign could be heading.

“We’re definitely expecting it to ramp up,” Le Roux told The New York Times. “I don’t think they’re going to stop. I think that we’re seeing the initial kickoff of the US election focused campaigning.”

That warning gives the developing story a much bigger significance than a handful of fake videos circulating online. If the activity does expand, voters could encounter an increasingly confusing mixture of genuine reporting, manipulated footage, fabricated news graphics and AI-generated material as the midterm campaign intensifies.

The alleged operation carries echoes of Russia’s intervention in the 2016 presidential election, when U.S. intelligence agencies concluded that Moscow conducted an influence campaign aimed at affecting the American political environment. Since then, previously declassified U.S. assessments have found Russian online influence activity during subsequent federal election cycles as well.

The technology, however, has changed dramatically.

Play

AI can now make fabricated material appear considerably more polished and believable, including videos that imitate the visual style of respected news organizations or manipulate the appearance and voices of recognizable public figures. In the latest activity, researchers identified videos making false or unsubstantiated claims about Democratic Senate candidates, including accusations involving corruption, antisemitism and transgender issues.

One example reportedly mimicked a New York Times-style video and used an image of actress Jamie Lee Curtis alongside subtitles criticizing Democrats. Curtis’s publicist rejected the material, saying there was “not even a scintilla of truth in that.” The fake video was viewed nearly 200,000 times on X before some related posts were removed.

The operation has been attributed by Graphika to a Kremlin-linked network known as Operation Overload, also called Matryoshka, which has previously been associated with impersonating legitimate media organizations and spreading fabricated content.

For Trump, the situation creates an unusually complicated backdrop for the midterms. The Russian campaign described by U.S. officials is not necessarily designed simply to help one American political party. Intelligence officials told The New York Times that the broader objective appears to be sowing division and chaos, although the recent examples identified by independent researchers have disproportionately targeted Democratic candidates.

That means the potential impact could extend beyond individual Senate races.

A campaign that successfully spreads fabricated stories can make voters question legitimate journalism, distrust political opponents and become uncertain about which information is authentic. The Brennan Center has separately warned that Russia, China and Iran are expected to use increasingly sophisticated AI capabilities in attempts to influence the 2026 midterms.

The most important warning may therefore be the simplest one: the first wave may not be the largest.

Le Roux’s assessment that researchers are seeing the “initial kickoff” suggests that the material already circulating could represent an early stage of a much broader campaign. Whether that happens remains to be seen, but U.S. intelligence officials and independent researchers are already watching for signs of escalation.

For Trump and the American political system, that is the ugly surprise emerging from the Kremlin’s latest playbook: Russia does not need to interfere with the machinery that counts the votes to create disruption. An increasingly sophisticated AI campaign can instead target the information surrounding those votes, flooding social media with material designed to make Americans question what they are seeing before they ever reach the ballot box.

And if Graphika’s warning proves correct, the videos appearing this week may be only the opening salvo.