'Citizens Against Mamdani' Accounts Reportedly Posted AI-Generated Videos of Fictional New Yorkers to Simulate Political Opposition

In New York, linked "Citizens Against Mamdani" accounts reportedly posted AI-generated videos of fictional constituents criticizing then-mayor-elect Zohran Mamdani across Instagram, TikTok, and X. Some videos reportedly displayed a visible Sora watermark, and forensic reviewers reportedly assessed at least one clip as highly likely to be a deepfake. The campaign reportedly appeared to simulate grassroots opposition and drew substantial engagement.

Risk classification

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

The incident involves a coordinated campaign using AI-generated deepfakes to simulate grassroots political opposition and manipulate public opinion.

Additional risk subdomains

• 3.2 Pollution of information ecosystem and loss of consensus reality: The automated astroturfing creates a 'mirage of political popularity' that pollutes the social media information ecosystem.

Causal factors

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

The campaign was intentionally launched by human actors using deployed AI models to generate political disinformation.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Risk Level 3. Limited Risk: The incident involves AI-generated content such as deepfakes. The report notes the presence of a 'Sora watermark' on some videos, which aligns with transparency obligations for synthetic media.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Deepfake Video Generation; Social Media Content Generation
• Behaviour type: Tool
• Alleged developer: Unknown voice cloning technology developers, Unknown deepfake technology developers, OpenAI
• Alleged deployer: Unknown social media account operators, Unknown disinformation actors targeting Zohran Mamdani, Unknown disinformation actors, Citizens Against Mamdani social media accounts, Citizens Against Mamdani
• Alleged harmed parties: Zohran Mamdani, Voters in New York, General public of New York, Epistemic integrity, Electoral integrity, American social media users

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 Negligible
• 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: Yes, the report describes the creation and spread of deepfake videos criticizing a politician.

Directly caused: The campaign 'Citizens Against Mamdani' created multiple deepfake videos of imaginary New Yorkers criticizing the mayor-elect, spreading them on TikTok, Twitter, and Instagram.

Inferred additional harm: It is likely that similar automated astroturfing campaigns will be deployed against other local politicians or businesses, spreading further malicious content.

Democracy

Reported: Yes, the report explicitly discusses how the campaign represents an automated form of astroturfing to manipulate political perception.

Directly caused: The deepfake campaign simulated grassroots opposition ('Citizens Against Mamdani') to influence public opinion against mayor-elect Zohran Mamdani.

Indirectly caused: The democratization of astroturfing via generative AI lowers the cost of political disinformation, potentially undermining trust in democratic elections.

Inferred additional harm: Widespread adoption of these techniques could severely erode public trust in local democratic processes and constituent representation.

Epistemic

Reported: Yes, the report describes the creation of fake citizens to generate a false impression of political opposition.

Directly caused: AI was used to generate hyperrealistic videos of non-existent New Yorkers with classic accents, creating a 'mirage of political popularity' or opposition.

Indirectly caused: The comments on the videos suggest some online users took the fake citizens and their complaints seriously, believing them to be real constituents.

Inferred additional harm: The proliferation of undetectable synthetic constituents could lead to a broader loss of shared consensus reality regarding public opinion.

People affected

• Occurrences reported: 1
• People reportedly harmed: 1
• People reportedly exposed: 30000

Potential causes

Technology

• Advanced Text-to-Video Tools: Generative AI like Sora makes hyperrealistic video creation cheap and easy.
• LLM Script Generation: LLMs allow rapid generation of diverse and convincing scripts at scale.
• Imperfect Detection Tools: No surefire way to confirm content is AI-generated without advanced tech.

Data Inputs

• Synthetic Persona Data: Algorithms use realistic accents and faces to mimic diverse populations.
• Unlabeled AI Content: Missing or easily ignored watermarks make identifying deepfakes hard.

Human Factors

• User Susceptibility to Deepfakes: Online users take fake videos seriously, believing the illusion of support.
• Low Barrier to Malicious Use: Only a few prompts are needed to generate political disinformation campaigns.

Process and Methods

• Automated Astroturfing: Combining LLMs and deepfakes automates the creation of fake support online.
• Inadequate Platform Moderation: Social media platforms host deepfakes without immediate detection or removal.

Information quality

• Classification confidence: High
• Reason for confidence: The report clearly details the nature of the campaign, the platforms used, the specific AI technologies involved (Sora, LLMs), and the assessments by cybersecurity firms (GetReal Security, Reality Defender) confirming the synthetic nature of the media.
• Ambiguities identified: The exact identity of the human creators behind the 'Citizens Against Mamdani' accounts remains unknown.
• Alternative interpretations: None. The evidence of AI generation (Sora watermark, 99% deepfake score) is overwhelming.

A coordinated digital influence campaign utilized generative AI tools, including OpenAI's Sora, to create deepfake videos of fictional New Yorkers criticizing NYC mayor-elect Zohran Mamdani. The incident represents an evolved capability in automated political astroturfing, though its national security impact remains minor and localized.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: United States
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. The incident represents an ongoing strategic and capability concern regarding AI-enabled political disinformation rather than an active, imminent national security crisis.
• Autonomy: Human-controlled. Human actors intentionally initiated the campaign, prompted the AI models (Sora and LLMs), and distributed the generated content across social media platforms.
• Novelty: Evolved capability. Represents an evolved capability where advanced text-to-video generators and LLMs are combined to lower the cost and increase the realism of automated political astroturfing.

Impact by dimension

• Physical security: Negligible. No physical systems, critical infrastructure, or kinetic assets were targeted or affected in this digital disinformation campaign.
• Information security: Minor. Involves AI-generated deepfake videos and automated astroturfing targeting a municipal mayor-elect. While representing information manipulation, the scale was limited and did not compromise national intelligence capabilities.
• Sovereignty: Minor. Targeted a local political figure (NYC mayor-elect) rather than national sovereign functions or federal electoral systems, presenting minor direct threat to state authority.
• Economic security: Negligible. No strategic technology theft, critical supply chain disruption, or financial system manipulation was reported in connection with this incident.
• Societal stability: Minor. The campaign attempted to manipulate public perception of a local politician, but did not cause mass civil unrest, systematic discrimination, or large-scale erosion of civil liberties.