This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
The Bangladesh Chief Adviser's Press Wing has issued a warning regarding the circulation of AI-generated deepfake videos on Facebook. These videos falsely depict government officials endorsing gambling applications like 'Blue Live' and 'Crazy Time Bangladesh' by manipulating footage from legitimate news interviews. The government has identified these as fraudulent attempts to lure the public into gambling sites and has urged citizens to exercise caution.
Bangladesh's Chief Adviser's Press Wing has warned the public about alleged AI-generated videos depicting Muhammad Yunus endorsing gambling apps. The videos reportedly reuse footage from legitimate interviews, manipulated with purportedly fabricated audio and visuals. At least two such alleged deepfakes have circulated on Facebook. The account that published them reportedly appears to be operated by multiple users located in Lithuania, misleading viewers with claims of guaranteed earnings. The government has urged users not to trust or share these videos.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
Using AI systems to gain a personal advantage over others such as through cheating, fraud, scams, blackmail or targeted manipulation of beliefs or behavior. Examples include AI-facilitated plagiarism for research or education, impersonating a trusted or fake individual for illegitimate financial benefit, or creating humiliating or sexual imagery.
Also classified as
The deepfakes generate fabricated statements and fake news reports, spreading misleading information to the public.
The incident involves malicious actors using AI-generated deepfakes of a public figure to deceive social media users for financial gain through fraudulent gambling apps.
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 gamblers intentionally creating and deploying deepfake videos to defraud the public using deployed AI tools.
Added categories peak at 3: Substantial (Epistemic Harm, Inferred)
AI-generated deepfake videos of Bangladesh's Chief Adviser were circulated on social media to promote fraudulent gambling applications. While the incident represents a minor threat to government trust and public information integrity, it is primarily a financially motivated scam rather than a coordinated national security threat.
The AI systems were used as tools by human creators to generate synthetic video and voice mimicking the Chief Adviser. The report describes the videos as 'AI-generated' and 'AI-manipulated' with a 'robotic' voice, indicating standard media generation tools.
Limited Risk: The report describes 'AI-generated deepfake videos' which fall under the transparency obligations of Risk Level 3 for AI-generated content.
The AI was used to generate deepfake videos and synthetic voices mimicking the Chief Adviser of Bangladesh to promote gambling apps.
The report clearly details the nature of the deepfakes, the platforms used, the specific apps promoted, and the government's response. The role of AI as a tool for malicious actors is explicit, and boundary rules clearly point to 4.3.
The incident could be viewed as a political disinformation campaign (4.1), but the primary intent is clearly financial fraud/scams targeting individuals, making 4.3 the most appropriate primary classification.
Technology
AI Deepfake Video Generation
AI tools were used to manipulate facial movements and mimic the Chief Adviser.
Robotic Voice Synthesis
AI voice cloning mimicked the victim's voice, though it remained robotic.
Data Inputs
Exploitation of Public Footage
Real interview footage from Al Jazeera and News24 was used as input data.
Unauthorized Use of Media Logos
The bdnews24.com logo was added to the video to fake credibility.
Human Factors
Targeting Vulnerable Users
Gamblers exploited public trust in the Chief Adviser to deceive citizens.
Malicious Actor Motivation
Lithuanian-operated accounts created deepfakes for financial gain.
Process and Methods
Inadequate Ad Moderation
Meta's platforms allowed fraudulent deepfake gambling ads to be distributed.
Weak Account Verification
Fake accounts created in 2022 could easily post deceptive content.
Management
Lack of Proactive Monitoring
Platform management failed to detect and block recurring deepfake campaigns.
Regulatory Environment
Weak Social Media Regulation
Lack of strict regulatory enforcement on deepfakes and illegal gambling ads.
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.
