This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video masquerading as a France 24 news report circulated on social media, falsely claiming that Ukrainian authorities planned an assassination attempt on French President Emmanuel Macron. The video used AI-generated voice cloning and manipulated visuals to impersonate a news anchor. France 24 confirmed the report was fake, and investigations linked the disinformation campaign to pro-Russian networks.
A purported deepfake video reportedly claimed France 24 reported a Kyiv plot to assassinate French President Macron. This reported fake news was debunked by France 24, which confirmed the video was altered and did not air any such report.
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 AI system was used to generate highly realistic but entirely fabricated news content to mislead the public.
The deepfake was created and distributed as part of a coordinated geopolitical disinformation campaign aimed at manipulating public opinion and undermining international relations.
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 incident was caused by human actors who intentionally used deployed AI deepfake and voice cloning tools to generate and spread geopolitical disinformation.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
A coordinated pro-Russian disinformation campaign used AI-generated voice cloning and deepfake technology to impersonate a France 24 news broadcast, falsely claiming an assassination plot against French President Emmanuel Macron. The incident highlights the growing national security threat of sophisticated, state-linked AI information operations aimed at undermining democratic alliances.
The AI systems (voice cloning and deepfake video generation software) were used as tools by human operators to perform specific, well-defined functions of generating synthetic media based on direct inputs.
Limited Risk: The incident involves the creation of AI-generated content, specifically a deepfake video and cloned voice, which falls under the transparency obligations of Risk Level 3.
The primary function of the AI was to generate a deepfake video of a news anchor, and the secondary function was to clone and generate his voice.
The reports from multiple independent fact-checking organizations and news outlets provide highly detailed, consistent, and verified accounts of the deepfake's creation, content, and dissemination.
None. The evidence that the video is a deepfake is clear and uncontested by all reputable sources.
Technology
AI Voice Cloning Software
Voice cloning technology used to mimic the presenter's voice.
Deepfake Video Generation
Neural networks used to alter the presenter's lip movements.
Digital Ticker Fabrication
On-screen graphics and news tickers were digitally manipulated.
Data Inputs
Harvesting Real Broadcasts
Original France 24 footage used as source data for deepfake.
Voice Sample Extraction
Real news bulletins used to extract voice data for cloning.
Human Factors
Malicious Intent of Creators
Propagandists actively sought to destabilize Ukraine and allies.
Amplification by Politicians
Politicians like Dmitry Medvedev shared the fake, increasing reach.
User Confirmation Bias
Social media users shared unverified content matching their views.
Process and Methods
Inadequate Content Filtering
Platforms failed to pre-screen and flag AI-generated deepfakes.
Reactive Fact-Checking
Fact-checks were published only after the video went viral.
Management
Coordinated Disinformation
Kremlin management utilized 'Portal Kombat' to direct propaganda.
Regulatory Environment
Weak Deepfake Regulation
Lack of international legal frameworks to deter deepfake creation.
Poor Platform Accountability
No regulatory pressure on platforms to immediately remove deepfakes.
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.
