Russian Disinformation Campaign Allegedly Used Fake News Site 'KBSF-San Francisco News' and Deepfake Video to Falsely Accuse Kamala Harris of 2011 Hit-and-Run

A Russian disinformation campaign allegedly used a fake news site, "KBSF-San Francisco News," and an AI-generated deepfake video to falsely accuse Kamala Harris of a 2011 hit-and-run accident in San Francisco. The deepfake featured a fabricated African-American victim, whose AI-generated testimony spread widely across social media. Microsoft identified the hoax as part of Russian cyber influence operations targeting the 2024 U.S. election, with the site registered in Iceland.

Risk classification

• Primary risk domain: 4 Malicious actors
• Primary risk subdomain: 4.1 Disinformation, surveillance, and influence at scale

The incident involves a coordinated, state-linked disinformation campaign using AI-generated deepfakes to influence the U.S. presidential election and manipulate public opinion at scale.

Causal factors

• Entity: Human
• Intent: Intentional
• Timing: Post-deployment

The incident was caused by human actors (the Russian disinformation group Storm-1516) intentionally deploying post-deployment generative AI tools to create deceptive deepfakes.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The report describes the creation of deepfakes ('AI-generated content such as deepfakes'), which falls under Risk Level 3 due to transparency obligations to inform users that the content is AI-generated.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Video Generation; Voice Generation
• Behaviour type: Tool
• Alleged developer: John Mark Dougan, Government of Russia
• Alleged deployer: Storm-1516, KBSF-San Francisco News, John Mark Dougan, Government of Russia
• Alleged harmed parties: National security and intelligence stakeholders, Media integrity, Kamala Harris, Journalistic integrity, Journalism, General public of the United States, General public, Epistemic integrity, Democracy

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Minor, indirect Severe
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Minor, indirect Minor
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Severe
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: Yes, the report explicitly describes the creation and spread of malicious deepfake videos and fake news articles.

Directly caused: The campaign directly created and spread a 4-minute AI-generated video of a fake hit-and-run victim, which was shared over 10,000 times on X within 24 hours.

Indirectly caused: The campaign led to the wider dissemination of other malicious deepfakes, including a video accusing Tim Walz of sexual assault that gained 5 million views on X.

Inferred additional harm: It is likely that millions of additional social media users were exposed to this malicious content across platforms like Facebook, TikTok, and YouTube where it was widely shared.

Democracy

Reported: Yes, the report explicitly describes foreign actors using AI-assisted disinformation to undermine faith in the democratic process and influence the U.S. election.

Directly caused: The campaign targeted U.S. political candidates, including Kamala Harris and Tim Walz, with highly realistic deepfakes to manipulate public opinion ahead of the election.

Indirectly caused: The operations aimed to cast doubt about the integrity of the election's outcome and undermine confidence in the democratic process.

Inferred additional harm: The widespread exposure to such deepfakes likely contributed to a broader erosion of trust in democratic institutions and electoral integrity among the electorate.

Epistemic

Reported: Yes, the report explicitly describes the fabrication of news, fake victim testimony, and manipulated media to construct a false narrative.

Directly caused: The campaign fabricated a 2011 hit-and-run accident using an AI-generated video of a non-existent victim and unrelated medical X-rays to deceive the public.

Indirectly caused: The spread of these highly realistic deepfakes across multiple platforms eroded the shared consensus reality regarding the candidates' backgrounds.

Inferred additional harm: The proliferation of such sophisticated deepfakes makes it increasingly difficult for the public to distinguish between real and fabricated news, leading to long-term epistemic degradation.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 5000000

Potential causes

Management

• State-Sponsored Orchestration: Foreign governments funded and directed systematic cyber influence campaigns.
• Coordinated Troll Operations: Groups like Storm-1516 managed multiple sites to amplify fake narratives.
• Campaign Security Failures: Inadequate email security allowed foreign spear-phishing to succeed.

Technology

• AI Deepfake Generation: AI was used to synthesize highly realistic voices and facial expressions.
• Social Media Platform Algorithms: Algorithms rapidly amplified deceptive videos to millions of viewers.
• Voice Manipulation Tools: TrueMedia identified a 97 percent probability of AI-generated audio.

Data Inputs

• Fabricated Victim Testimonies: Actors and AI-generated avatars were used as inputs for fake stories.
• Misappropriated Academic Images: Real orthopedic x-ray scans from 2010 and 2017 were used to fake injuries.
• Spoofed Media Brand Elements: Fake domains like KBSF-TV were registered to mimic local news outlets.

Human Factors

• Malicious State-Backed Actors: Russian, Iranian, and Chinese groups actively targeted US elections.
• Voter Confirmation Bias: Emotional narratives easily deceived users into sharing fake news.
• Lack of Source Verification: Viewers shared deepfakes without verifying facts with credible outlets.

Process and Methods

• Inadequate Content Moderation: Platforms failed to block fake videos before they gained millions of views.
• Easy Domain Registration: Fake news domains were easily registered anonymously in foreign countries.
• State-Level Leak Campaigns: Hack-and-leak operations bypassed traditional campaign security.

Regulatory Environment

• Jurisdictional Escape: Malicious actors fled to Russia to evade US criminal investigations.
• Weak Platform Accountability: Lack of regulations allowed platforms to host unverified election deepfakes.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed accounts from credible sources like Microsoft Threat Analysis Center and TrueMedia regarding the nature of the deepfakes, the platforms involved, and the specific actors (Storm-1516) behind the campaign.
• Ambiguities identified: The specific AI tools or models used to generate the deepfakes are not identified in the reports.
• Alternative interpretations: None. The evidence clearly points to a coordinated, politically motivated disinformation campaign using synthetic media.

A Russian-linked group deployed sophisticated AI-generated deepfakes to falsely accuse U.S. Vice President Kamala Harris of a hit-and-run. This incident represents a severe information warfare and sovereignty threat, leveraging synthetic media to actively manipulate the democratic electoral process of a sovereign nation.

• Overall national security impact: Severe
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Storm-1516

Threat characteristics

• Imminence: Near-term. The campaign was active and developing in the lead-up to the 2024 presidential election, requiring ongoing monitoring.
• Autonomy: Human-controlled. AI tools were used strictly as instruments by human operators to generate specific synthetic media assets.
• Novelty: Evolved capability. Represents a significant advancement in the sophistication and deployment scale of AI-generated deepfakes for political influence.

Impact by dimension

• Physical security: Negligible. The incident involved no physical threats, kinetic targeting, or critical infrastructure manipulation.
• Information security: Severe. Coordinated foreign state-linked disinformation campaign utilizing sophisticated AI deepfakes to target high-level political leadership and manipulate public opinion.
• Sovereignty: Severe. Direct attempt by a foreign adversary to manipulate the democratic electoral process and influence the outcome of the U.S. presidential election.
• Economic security: Negligible. No financial systems, strategic technology theft, or critical supply chains were targeted or affected in this incident.
• Societal stability: Substantial. The campaign targeted public trust, aiming to erode confidence in democratic institutions and political candidates at a national scale.