This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Foreign state actors, including those from Russia and China, are utilizing AI-generated content to manufacture and amplify opposition to data center construction projects in the U.S. and Armenia. These campaigns involve fabricated news articles, AI-generated imagery, and automated social media accounts designed to exploit local grievances and erode trust in infrastructure and Western technology. While the impact on public opinion is still being assessed, these efforts represent a coordinated attempt to deepen domestic fractures and undermine technological development.
Alethea reported that Storm-1516, a Russian influence operation, targeted Firebird's NVIDIA-powered data center under construction in Hrazdan, Armenia, with purportedly fabricated media narratives designed to make the project appear dangerous and economically unstable. The campaign allegedly used imitation tech-news articles to cast the facility as a liability for Armenia's infrastructure and Western-aligned technology ambitions.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to conduct large-scale disinformation campaigns, malicious surveillance, or targeted and sophisticated automated censorship and propaganda, with the aim of manipulating political processes, public opinion, and behavior.
Also classified as
The high-volume generation of AI 'slop' and fabricated news articles pollutes the information ecosystem and degrades the quality of public discourse.
Foreign state actors are using AI-generated content and social media campaigns to conduct coordinated, large-scale influence operations aimed at manipulating U.S. public opinion and political debate surrounding data centers.
Human
Due to a decision or action made by humans
Intentional
Due to an expected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The risk is driven by human adversaries who intentionally use deployed AI systems to generate and spread divisive propaganda.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
Foreign state actors from China, Russia, and Iran are using generative AI to manufacture and scale disinformation campaigns targeting U.S. and Armenian data center projects. By exploiting local infrastructure disputes, these adversaries aim to undermine Western technological development and degrade public trust. While direct impact remains limited, it represents a substantial information warfare and economic security concern.
The AI systems, such as ChatGPT, were used by human operatives to generate text, comic strips, and video scripts. The AI acted as an assistant to augment the creative and linguistic capabilities of the foreign actors, requiring human prompting and deployment to execute the campaigns.
Risk Level 3: Limited Risk. The report describes the use of chatbots like ChatGPT and AI-generated content (images and narrated videos) to create deepfakes and synthetic media, which are subject to transparency obligations under the EU AI Act.
The AI systems were primarily used by operatives to generate social media posts, comic strips, and scripts for coordinated online campaigns.
The reports provide high-quality, detailed threat intelligence from credible sources (The New York Times and Alethea) with specific examples of campaigns, platforms used, and attribution details. The role of AI systems like ChatGPT is explicitly stated and confirmed by OpenAI, leaving little ambiguity about the nature of the incident.
Some actors argue that the opposition to data centers is entirely organic and that foreign interference is being exaggerated by industry lobbyists to discredit legitimate environmental concerns.
Technology
Abuse of GenAI Tools
Operatives used ChatGPT to generate deceptive social media campaigns.
AI-Generated Slop and Images
Anonymously run pages used AI to scale emotional, anti-data center imagery.
Fake News Generation
Russian Storm-1516 used AI to generate fake tech journalist articles.
Data Inputs
Thermal and Satellite Data
Thermal and satellite imagery was used to exaggerate environmental impacts.
Inauthentic Location Data
Accounts run from Bangladesh used fake local US names to mimic residents.
Human Factors
Public Anxiety and Hostility
Deep public ambivalence and fear about AI infrastructure were easily exploited.
Susceptibility to Rage Bait
Users engaged with emotionally charged, AI-generated 'rural rage bait'.
Adoption of Adversary Tactics
Domestic political candidates copied foreign AI-generated video aesthetics.
Process and Methods
Inadequate Platform Moderation
Social media platforms failed to quickly block AI-generated campaigns.
Lack of Attribution Capabilities
Difficulty in identifying the source of AI-generated content in real time.
Absence of Joint Threat Response
Security and communication teams worked in silos rather than a unified front.
Management
Misreading Narrative Threats
Failure to recognize coordinated narrative campaigns as security reconnaissance.
Siloed Threat Intelligence
Management did not integrate CISO, CCO, and government affairs response.
Slow Response Windows
Organizations lacked proactive monitoring to catch fake claims before escalation.
Regulatory Environment
Lack of AI Content Standards
Absence of regulations on labeling AI-generated political content.
Weak Foreign Influence Oversight
Dismantling of government teams that tracked foreign influence operations.
Permit and Siting Vulnerabilities
Fragmented local regulatory processes allowed campaigns to stall infrastructure.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
