This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video featuring a Burkina Faso military officer was created for entertainment purposes but was widely circulated on social media platforms as an authentic political speech. Fact-checkers confirmed the video was AI-generated using detection tools from Hive Moderation and Sensity. Despite the original creator including a disclaimer that the content was fictional, the video was misinterpreted by the public as real.
A video reportedly published on May 25, 2025, claimed to show Burkina Faso President Ibrahim Traoré delivering a forceful speech against Western media and exploitation. The video, which purportedly used AI-generated visuals and audio, was shared widely on social platforms. Despite containing disclaimers identifying it as fictional, the video was reportedly circulated and interpreted by some viewers as authentic.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that inadvertently generate or spread incorrect or deceptive information, which can lead to inaccurate beliefs in users and undermine their autonomy. Humans that make decisions based on false beliefs can experience physical, emotional or material harms
Also classified as
The deepfake was circulated widely on social media platforms to influence public opinion regarding Western exploitation of Africa.
The incident involves an AI-generated deepfake video that was circulated as a real speech, leading to the spread of false and misleading political information among social media users.
Human
Due to a decision or action made by humans
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed
The risk occurred due to an unexpected outcome from pursuing a goal, as the creator intended the video for entertainment but humans circulated it as real.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
An AI-generated deepfake of a Burkina Faso military officer was circulated as an authentic political speech. While the video spread political misinformation on social media, the incident represents a minor national security impact, manageable within standard security and fact-checking procedures, with no evidence of coordinated state-sponsored information warfare.
The AI systems involved (both the unnamed deepfake generator and the Hive Moderation detection tool) operate as tools. They perform specific, well-defined functions (generating synthetic media or detecting it) based on direct inputs without autonomous planning.
Limited Risk: The incident involves AI-generated content (a deepfake video), which falls under Risk Level 3 due to transparency obligations for synthetic media.
The primary AI system was used to generate a synthetic deepfake video of a military officer, while the secondary AI system (Hive Moderation) was used for image and audio verification to detect the deepfake.
The report clearly details the nature of the deepfake, its origin (created for entertainment with a disclaimer), and its subsequent circulation as real. Fact-checking and AI detection probabilities are explicitly provided.
None. It is clearly a case of a synthetic video being circulated as authentic political speech.
Technology
High-Fidelity AI Synthesis
AI tools generated highly realistic video and audio of the leader.
Absence of Automated Watermarking
The synthetic content lacked embedded, unalterable metadata or watermarks.
Data Inputs
Biometric Data Exploitation
Real footage and voice recordings were used to synthesize the deepfake.
Human Factors
Deceptive clipping by users
Bad actors clipped a 15-second segment to remove the original context.
Viewer gullibility
Social media users believed and shared the deepfake as authentic.
Process and Methods
Ineffective disclaimer preservation
The fictional disclaimer was in the description, not embedded in the video.
Lax platform detection protocols
Platforms failed to automatically scan and flag the uploaded synthetic clip.
Management
Inadequate risk mitigation
The creator relied on easily removable text disclaimers for safety.
Regulatory Environment
Lack of synthetic media standards
No legal frameworks mandated platforms to block or tag unlabeled deepfakes.
AI system involved
Identified by automated extraction; can differ from the alleged developers above.
