American Asylum Seeker John Mark Dougan in Russia Reportedly Spreads Disinformation via AI Tools and Fake News Network

John Mark Dougan, a former Florida sheriff's deputy granted asylum in Russia, has been implicated in spreading disinformation. Utilizing AI tools like OpenAI's ChatGPT and DALL-E 3, Dougan created over 160 fake news sites, disseminating false narratives to millions worldwide. His actions align with Russian disinformation strategies targeting Western democracies. See also Incident 734.

Risk classification

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

The incident describes a coordinated, state-linked influence operation using AI to generate and scale disinformation targeting political processes and public opinion across multiple countries.

Additional risk subdomains

• 3.2 Pollution of information ecosystem and loss of consensus reality: The network of over 160 fake news sites flooded the information ecosystem with AI-generated articles, which were then ingested and repeated by major commercial chatbots.
• 1.2 Exposure to toxic content: The network generated and spread highly targeted, defamatory, and malicious deepfakes and articles, such as false allegations of sexual misconduct targeting political candidates.

Causal factors

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

The incident was caused by human actors (John Mark Dougan and Russian state-sponsored operatives) intentionally using deployed commercial AI models to generate and scale disinformation.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The incident involves the use of AI systems to generate text and deepfakes, which are subject to transparency obligations under the EU AI Act to ensure users are aware they are interacting with AI-generated content.

AI system and alleged parties

• AI system: ChatGPT, DALL-E (OpenAI)
• AI purpose: Social Media Content Generation; Writing Assistant
• Behaviour type: Tool
• Alleged developer: OpenAI
• Alleged deployer: John Mark Dougan
• Alleged harmed parties: Journalism, Information integrity, General public, American citizens

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 Minor
• 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 Minor
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Malicious content

Reported: The report explicitly describes malicious and defamatory content created and spread directly by the incident.

Directly caused: The network created AI-generated deepfake videos, fabricated invoices, and hundreds of articles containing false allegations of corruption and sexual misconduct.

Indirectly caused: The malicious content was amplified by pro-Kremlin influencers, botnets, and social media users, reaching over 12 million accounts on platforms like X.

Inferred additional harm: It is highly likely that millions of additional social media users were exposed to the toxic and polarizing content across various platforms.

Democracy

Reported: The report explicitly describes efforts to erode democratic norms and undermine faith in the integrity of the democratic process.

Directly caused: The AI-scaled disinformation campaigns targeted the 2024 U.S. and German elections, attempting to shape voter preferences and sow distrust in public institutions.

Indirectly caused: The amplification of these narratives by political figures and mainstream media outlets threatened to legitimize foreign propaganda within domestic political discourse.

Inferred additional harm: The cumulative effect of persistent foreign influence operations likely contributes to long-term polarization and a decline in public trust in democratic elections.

Epistemic

Reported: The report explicitly describes extensive epistemic harm, including the generation of false narratives and the corruption of the information environment.

Directly caused: The network generated hundreds of fake news articles and deepfakes that successfully deceived millions of social media users and were indexed as top search results on Google.

Indirectly caused: Ten major commercial AI chatbots repeated the Russian disinformation narratives 32 percent of the time, validating and further spreading the falsehoods to users seeking accurate information.

Inferred additional harm: The widespread integration of fake news sites into search engines and chatbot training data causes ongoing, unquantified degradation of the shared information ecosystem.

People affected

• Occurrences reported: 1
• People reportedly harmed: 5
• People reportedly exposed: 12000000

Potential causes

Management

• Inadequate Platform Verification: Tech platforms failed to proactively prevent mass creation of fake accounts.
• Delayed Action on Fake Domains: Adversaries easily recreated seized domains within 24 hours under new TLDs.
• Prioritizing Outreach Over Safety: AI companies deployed models without ensuring immunity to foreign propaganda.

Technology

• Generative AI Content Scaling: Commercial AI tools enabled rapid scaling of low-quality propaganda articles.
• AI Voice Cloning and Deepfakes: AI cloned the FBI director's voice and created fake videos to spread fraud claims.
• Algorithmic Prioritization of Engagement: Social media algorithms prioritized high-engagement extreme content over truth.

Data Inputs

• Chatbot Training on Propaganda Sites: Chatbots ingested fake local news sites, repeating Russian narratives 32 percent of the time.
• Lack of Source Reliability Filtering: AI models failed to recognize propaganda front sites like 'The Boston Times'.
• Plagiarized and Rewritten Content: Generative AI was used to rewrite plagiarized articles from mainstream media.

Human Factors

• User Susceptibility to Sensationalism: Audiences easily shared shocking deepfakes and stories without verification.
• Influencer Amplification of Falsehoods: High-profile influencers shared fake narratives, exposing them to millions.
• Trust in Chatbots for Accurate Info: Users increasingly relied on AI chatbots for election-related information.

Process and Methods

• Narrative Laundering Techniques: Using paid actors and fake sites to make disinformation appear as legitimate news.
• Inadequate Chatbot Guardrails: AI models lacked robust filters to prevent repeating debunked conspiracy theories.
• Evasion of Platform Moderation: Operations used redirect chains and cloaking to bypass social media detection.

Regulatory Environment

• Lack of Global AI Safety Standards: Absence of enforceable international regulations on generative AI deployment.
• Permissive Hosting Environments: Bulletproof hosting providers allowed malicious sites to remain online.
• Weak Know-Your-Customer Requirements: Hosting services permitted purchasing infrastructure without identity verification.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide detailed, consistent, and well-sourced accounts from multiple independent organizations (NewsGuard, BBC Verify, Microsoft, Clemson University) detailing the operation, the specific AI tools used, and the reach of the disinformation campaigns.

A massive, Russian GRU-linked disinformation network led by a US fugitive utilized commercial generative AI tools to scale influence operations targeting US and European elections. By generating deepfakes and automating articles across 160 fake news sites, the operation successfully spread propaganda to millions, demonstrating how AI lowers the cost and increases the reach of state-sponsored information warfare.

• Overall national security impact: Severe
• Scope: Multiple nations
• Primary target: United States
• Other affected: Germany, Ukraine
• Alleged perpetrator: John Mark Dougan and the Russian GRU

Threat characteristics

• Imminence: Near-term. The operation represents an active, ongoing influence campaign targeting upcoming and active election cycles, requiring coordinated monitoring and defensive action.
• Autonomy: Human-controlled. The generative AI tools (ChatGPT and DALL-E 3) were used directly as tools by human operators to generate specific text and images, rather than acting on their own initiative.
• Novelty: Evolved capability. While foreign election interference and fake news sites are established threats, the use of commercial generative AI to mass-produce realistic content and deepfakes at this scale represents a significant evolution in capability.

Impact by dimension

• Physical security: Negligible. The incident was strictly limited to digital disinformation and did not involve kinetic attacks, targeting systems, or manipulation of physical critical infrastructure.
• Information security: Severe. A large-scale, state-linked (GRU) disinformation network used generative AI to scale influence operations, creating over 160 fake news sites. They successfully spread fabricated narratives, deepfakes, and propaganda that reached millions of viewers, directly compromising the integrity of the information ecosystem during key political periods.
• Sovereignty: Severe. The Russian-linked AI network actively targeted the democratic electoral processes of the United States and Germany, spreading fabricated narratives about political candidates and foreign policy to manipulate voter preferences and erode public trust in government institutions.
• Economic security: Minor. The incident demonstrates a vulnerability where major commercial AI models (chatbots) repeatedly ingested and validated foreign propaganda, showing a technological gap in filtering state-sponsored disinformation.
• Societal stability: Substantial. The campaign systematically targeted political figures and institutions with highly polarized, defamatory narratives, aiming to exacerbate social division, undermine democratic norms, and erode public trust.