Artificial intelligence news.
Straight from the original announcements.

Russian Network Used Deepfake Abuse Claim Against Texas Senate Candidate

October 23, 2026By the AI News Bites Editorial TeamSource: Resemble AI
Russian Network Used Deepfake Abuse Claim Against Texas Senate Candidate
Image: Resemble AI

Two videos posted to X on October 20 that showed a former student accusing Texas Democratic Senate candidate James Talarico of sexual abuse were AI deepfakes tied to the Russian influence network Storm-1516, according to a PolitiFact fact-check cited in Resemble AI's weekly Deepfake Watchlist. The videos were built around a real person who says he never made the accusation.

Key points
  • PolitiFact confirmed two X videos accusing James Talarico of abuse were deepfakes tied to Russia's Storm-1516.
  • The real person depicted, Zyon Lewis, says he never made the accusation; X suspended the account but reshares stayed up.
  • An Alameda County judge dismissed Mendones v. Cushman & Wakefield over an AI-generated witness video, per NBC News.
  • An 88-year-old lost $450,000 to a Mark Carney deepfake investment scam, according to CBC Go Public.
  • 34 anonymous TikTok accounts targeted Hungarian opposition leader Péter Magyar, per EDMO reporting.

A real person's identity used as the vehicle

Resemble AI's Deepfake Watchlist for the week of October 16 to 22, 2026, features the Talarico case as its top incident. PolitiFact confirmed on October 22 that the videos were deepfakes linked to Storm-1516. Darren Linvill of Clemson University's Media Forensics Hub confirmed the attribution, and Graphika found the posting account "very likely" impersonated a real Facebook user.

The person the video claimed to show, Zyon Lewis, told Lead Stories the video is not him and that he has made no such accusation. According to the watchlist, the account was created in September 2026 and registered in Romania. X suspended it by October 22, but reshares of the video remained live.

Resemble AI notes that no federal law specifically criminalizes AI-generated political defamation.

Forensic signs were visible without special tools

Fact-checkers and researchers pointed to several signatures in the videos:

  • Unnatural eye movement and blinking inconsistent with live speech
  • Stiff, repetitive facial expressions
  • A 98% AI-likelihood score from an AI detection tool
  • Teeth that differ from those in the real Lewis's Facebook photographs, suggesting a composite face
  • An audible African accent, which did not fit a person who attended a Texas school in 2012
  • British spellings in the on-screen caption

Resemble AI argues the effectiveness came less from technical sophistication than from anchoring the fabrication to a real person and from timing, two weeks before Election Day. It says the harm outlasts removal, because Lewis's name is now tied through search to an accusation he did not make. Linvill noted Russia likely aims to deepen partisan division and encourage turnout for Talarico's opponent.

Court evidence, a costly scam and a Hungarian campaign

The watchlist also covers three other cases. NBC News reported on October 20 that Alameda County Judge Victoria Kolakowski dismissed Mendones v. Cushman & Wakefield on September 9 after finding an AI-generated witness video in the plaintiffs' evidence. It may be the first civil case dismissed specifically over AI-fabricated legal evidence. The judge noticed a monotone voice, a slightly blurred face, repetitive movement and metadata indicating an iPhone 6 where the claims required an iPhone 15 or newer. Resemble AI says no equivalent federal rule on deepfake evidence exists, and the U.S. Judicial Conference declined in May to adopt a deepfake-specific authenticity standard.

CBC Go Public reported that Ralph Moore, an 88-year-old from Qualicum Beach, BC, sent $450,000 over ten months after seeing an AI-generated video of Prime Minister Mark Carney promoting a fraudulent investment platform. TD Bank questioned transactions, warned Moore and his wife, flagged the account and temporarily restricted it. The account was closed in February, and the bank has not reimbursed him.

European Digital Media Observatory reporting documented 34 anonymous TikTok accounts using deepfakes and fabricated news content against Péter Magyar, the leading opposition figure in Hungary's upcoming elections. Resemble AI says TikTok removed some accounts and that Hungary has no AI election content disclosure law.

What it means

Resemble AI's reading is that the incidents share a pattern: deepfakes aimed at moments and settings where trust is high or scrutiny is low, such as the closing weeks of an election, a courtroom or a long-running investment relationship. In several cases the company points to gaps in law or procedure rather than failures of detection, since the telltale signs in the Talarico and court cases were spotted without specialized tools.

Related news

Resemble AI tracks AI-cloned CEOs, battlefield fakes and voice swatting
Safety and ethicsNovember 12, 2026

Resemble AI tracks AI-cloned CEOs, battlefield fakes and voice swatting

Resemble AI's latest Deepfake Watchlist details a week of cases involving AI-cloned executives, manipulated battlefield videos, lawsuits over generated music, synthetic…

Deepfake Watchlist tracks election attacks, abusive apps and crypto fraud
Safety and ethicsNovember 5, 2026

Deepfake Watchlist tracks election attacks, abusive apps and crypto fraud

Resemble AI's latest Deepfake Watchlist examines three incidents involving altered news footage, concealed sexual deepfake tools and an impersonation-based cyberattack…

Resemble AI launches native real-time deepfake detection in Microsoft Teams
Safety and ethicsNovember 3, 2026

Resemble AI launches native real-time deepfake detection in Microsoft Teams

Resemble AI has launched a native real-time deepfake detection integration for Microsoft Teams. It scans audio and video during meetings for synthetic media, identity…