This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
In late December 2025, a series of purportedly AI-generated videos surfaced on TikTok advocating for Poland's exit from the European Union. Polish officials identified these videos as a potential disinformation campaign designed to test political narratives ahead of the 2027 parliamentary elections. The Polish government has requested that the European Commission investigate TikTok's compliance with the Digital Services Act regarding the spread of this content.
Purportedly AI-generated TikTok videos depicting young women in patriotic dress reportedly urged Poland to leave the EU ("Polexit") and spread widely in late December 2025. Poland's junior digital minister reportedly said the wave appeared to test narratives ahead of the 2027 election, and asked the European Commission to examine TikTok's DSA compliance. TikTok reportedly said it removed content that violated its rules.
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 campaign utilized AI-generated videos to spread false or misleading political narratives about Poland leaving the EU.
The incident involves a coordinated campaign using AI-generated videos to spread political disinformation and influence public opinion ahead of Poland's 2027 elections.
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 was caused by human actors who intentionally created and deployed AI-generated videos to conduct a political disinformation campaign.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
In late December 2025, a suspected foreign-aligned disinformation campaign utilized AI-generated TikTok videos to promote Poland's exit from the EU. Described by Polish officials as a test of political narratives ahead of the 2027 elections, the incident highlights the evolving threat of synthetic media in foreign influence operations targeting democratic processes.
The AI system was used as a tool by human actors to generate synthetic video content (deepfakes) for a political campaign, without exhibiting independent agency or autonomous decision-making.
Limited Risk: The incident involves AI-generated deepfakes, which under the EU AI Act require transparency obligations to inform users that the content is synthetic.
The AI system was used to generate synthetic videos of young women delivering political messages to be shared on TikTok.
The report clearly describes the political context and the government's response, but the specific AI tools used and the identity of the actors behind the campaign remain unconfirmed.
The videos could be interpreted as organic political expression rather than a coordinated foreign disinformation campaign, though government officials suspect the latter.
The specific AI models used to generate the videos, the identity of the creators, and quantitative data on the reach and engagement of the videos on TikTok.
Technology
AI Video Generation Tools
Allowed easy creation of realistic deepfake-style videos of patriotic women.
Recommendation Algorithm Exploitation
Algorithms prioritized sensational political content, aiding rapid spread.
Data Inputs
Lack of Provenance Metadata
Uploaded videos lacked metadata verifying AI origin or synthetic nature.
Targeted Narrative Data
Used divisive local political themes to maximize audience engagement.
Human Factors
User Susceptibility
Viewers easily influenced by realistic AI-generated patriotic personas.
Malicious Actor Intent
Foreign actors sought to test narratives and divide Polish society.
Process and Methods
Inadequate Proactive Detection
TikTok lacked automated tools to flag synthetic political disinformation early.
Reactive Moderation Process
Content was only removed after government escalation rather than proactively.
Management
Inadequate Risk Assessment
TikTok failed to anticipate AI-driven election interference on its platform.
Slow Root Cause Investigation
Platform struggled to trace the origin and creators of the AI campaign.
Regulatory Environment
Lagging DSA Enforcement
DSA compliance frameworks were not yet fully preventing such campaigns.
Lack of AI Disinformation Laws
Absence of strict legal penalties for deploying synthetic political campaigns.
AI system involved
An AI system was involved, but the reports do not identify it.
Identified by automated extraction; can differ from the alleged developers above.
