For Weeks, Deepfake Celebrities Have Been Telling Americans Not To Vote For Democrats In The Midterm Elections. U.S. Intelligence Says Russia Is Behind It.

TL;DR

Before you orderOffer from Amazon

Get movie nights delivered free with Prime

• Fast, free delivery on millions of items
• Prime Video, Amazon Music and more included
• Member-only deals all year

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

Sen. Mark Warner and Rep. Jim Himes say a classified U.S. intelligence report attributes an online campaign against congressional candidates to Russia. The campaign has included AI-generated celebrity videos urging people not to vote for Democrats, while officials have found no proven Russian attempts to hack election systems or campaign networks.

U.S. intelligence attributes an online campaign targeting congressional candidates to Russia, according to Sen. Mark Warner and Rep. Jim Himes, who spoke to Politico. The campaign includes AI-generated videos of celebrities urging Americans not to vote for Democrats ahead of the November 3 midterm elections; officials have found no proven Russian attempts to hack election systems or campaign networks.

Warner and Himes, both Democrats serving on their chambers’ intelligence committees, said their assessment is based on a classified intelligence report they received earlier in the week. The report says Moscow is using artificial intelligence tools to spread disinformation and weaken trust in individual candidates and in the elections overall. The lawmakers said Russia could target Democrats and Republicans who support Ukraine.

Warnings about election-related interference surfaced in early September. The Bot Blocker project, which monitors pro-Russian bot activity on social media, identified AI-generated videos featuring deepfakes of Robert De Niro, Julia Roberts and Cher telling viewers not to vote for Democrats. The project said a Russian bot network known as Matryoshka circulated the clips on X, using the logos of real news outlets and making false accusations against U.S. politicians.

A separate set of videos targeted Texas Democrat James Talarico. Researchers from Graphika, Alethea and Clemson University identified a network behind the material, NBC News reported. On October 20, posts on X made baseless allegations that Talarico had committed sexual assault; one post received more than 1.3 million views. The researchers identified the operation as Storm-1516, which has been linked to Russian military intelligence. These attributions come from researchers and reporting, not from publicly released details of the classified report.

How the Videos Could Affect Voters

The reported campaign seeks to influence voters without having to penetrate election infrastructure. False videos and fabricated accusations can circulate as political content, potentially undermining confidence in candidates and the voting process even when the underlying claims are untrue. That makes online distribution and credibility central to the concern raised by lawmakers.

The distinction between influence operations and hacking matters. The intelligence assessment described by Warner and Himes points to disinformation activity, while the agencies have found no proven efforts to breach voting systems or campaign networks. The absence of identified hacking attempts does not settle what impact the videos may have had on voters, or whether further activity is planned.

Researchers have also warned that large view totals can be misleading. Bloomberg reported that posts from Russian bot accounts on X can collect tens of thousands to hundreds of thousands of views through fake engagement. Some clips reportedly reach as many as 200,000 views in an hour, then stop gaining views and receive very few likes or reposts. Those figures describe reported platform activity, not verified exposure to that many individual voters.

Warnings Began in September

Reports of attempts to interfere in the congressional elections first surfaced in early September, when Bot Blocker flagged celebrity deepfakes circulating online. The clips’ use of familiar faces and real news organizations’ branding could make fabricated political messages appear more credible to viewers, although the source material does not establish how many people believed them.

The later Talarico videos illustrate a separate reported strand of activity: fabricated allegations directed at a named candidate, rather than a general appeal to avoid voting for one party. Researchers have associated the network behind those posts with Storm-1516, an operation previously accused of interfering in elections in the United States, Germany, Hungary and France. Those earlier accusations are part of the reported background; they do not by themselves establish responsibility for each new post.

“Moscow is using AI tools to spread disinformation and undermine trust in individual candidates and in the elections as a whole.”

— The U.S. intelligence report, as described by Warner and Himes

Attribution and Reach Remain Unclear

The classified report has not been publicly released in the source material, so its evidence, scope and level of confidence cannot be independently examined here. The lawmakers’ statements provide an account of its conclusions, but do not disclose which specific videos or accounts the report assessed.

It is also unclear how many authentic users saw the clips, how many believed them, and whether the posts changed anyone’s voting intentions. Reported view totals may be inflated by automated or fake engagement, and views do not establish that content was watched in full. The supplied reporting identifies no proven Russian attempt to hack election infrastructure or campaign networks, but that finding does not answer whether other forms of interference may emerge before Election Day.

Monitoring Through Election Day

The midterm elections are scheduled for November 3, leaving a short period in which further posts or new allegations could appear. The source material does not specify any announced investigation, platform enforcement action or public release of the intelligence report. Readers should distinguish official statements, researchers’ attributions and the videos’ false allegations rather than treating the clips as authentic evidence.

Further clarity would depend on whether U.S. officials release more information about the campaign and whether researchers or platforms document additional activity. Until then, the confirmed public account is limited: lawmakers say intelligence links Russia to an AI disinformation effort, researchers have attributed particular networks and posts, and the reported impact on voters remains unknown.

Key Questions

What does the U.S. intelligence assessment say?

Sen. Mark Warner and Rep. Jim Himes told Politico that a classified report says Russia is using AI tools to spread disinformation and erode trust in candidates and the elections. The report itself has not been publicly provided in the source material.

What do the celebrity videos show?

Bot Blocker reported AI-generated deepfakes of Robert De Niro, Julia Roberts and Cher urging viewers not to vote for Democrats. The videos were described as fabricated and circulated on X with real news outlets’ logos.

Were voting systems or campaign networks hacked?

The intelligence agencies have found no proven Russian attempts to hack election systems or campaign networks, according to the reporting. The reported activity concerns online disinformation, not a confirmed breach of voting infrastructure.

How much did the videos influence voters?

That is not known. Reported view counts may include fake engagement and do not show whether individual people watched or believed the videos, or whether their voting decisions changed.

When are the midterm elections?

The U.S. midterm elections are scheduled for November 3, 2026. The reported online campaign was continuing in the weeks before the vote.

Source: rss