This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video and associated fake news articles were used to impersonate the Sri Lankan President, Anura Kumara Dissanayake, to promote a fraudulent government investment scheme. The campaign used manipulated footage and AI-generated voice cloning to deceive citizens into clicking phishing links. Authorities, including the Presidential Media Division and Sri Lanka CERT, confirmed the content was AI-generated and part of a broader scam operation.
A reported deepfake depicts Sri Lankan President Anura Kumara Dissanayake endorsing a government investment plan. It reportedly uses manipulated footage from a May 3rd Satana TV program, dubbed or voice-cloned into English, and including a news article imitating the Daily Mirror to lure people to sign up for the scheme. Sri Lanka CERT, journalists, and the Presidential Media Division reportedly confirmed the video and associated links were AI-generated and part of a phishing and scam operation.
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 scam utilized phishing links designed to compromise users' personal details and social media credentials.
The deepfake video and fake news articles generated and spread false information about a government investment scheme.
The incident involves the use of AI-generated deepfakes to impersonate a public figure for the purpose of executing a financial scam and phishing campaign.
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 scammers who intentionally deployed AI voice cloning and video manipulation tools to defraud the public.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
An AI-generated deepfake and cloned voice of Sri Lankan President Anura Kumara Dissanayake were used to promote a fraudulent government investment scheme. While the incident impersonated a head of state, its primary objective was financial fraud and credential harvesting rather than political subversion, representing a minor national security impact that was managed through standard procedures.
The AI was used as a tool to generate synthetic voice and manipulate video footage based on direct inputs from the scammers. The report states the video was 'created using artificial intelligence by manipulating real footage' and 'dubbed or AI-generated voice', indicating standard media generation tool behavior.
Limited Risk: The report describes the use of AI-generated content, specifically deepfakes ('This video has been created using artificial intelligence by manipulating real footage'). Under the EU AI Act, deepfakes and AI-generated content are classified as limited risk, requiring transparency obligations.
The AI was used to manipulate real video footage of the President and clone his voice to create a realistic deepfake video for social media.
The fact-checking report is highly detailed regarding the nature of the scam, the specific media manipulated, and the verification steps taken with government and media officials. There is no ambiguity that the video is an AI-generated deepfake used for a financial scam.
None
Technology
AI Voice Cloning
AI was used to clone the President's voice with unnatural tone.
Deepfake Video Generation
Real video footage was manipulated using AI to match the fake audio.
Data Inputs
Unauthorized Use of Real Footage
Real TV broadcast footage was used as input for the deepfake.
Human Factors
Public Susceptibility to Scams
Users trust fake celebrity endorsements and high return claims.
Lack of Verification by Users
Social media users click external links without verifying sources.
Process and Methods
Lack of Deepfake Detection
Social media platforms failed to detect and flag the AI video.
Phishing Redirection Mechanisms
Scammers used fake URLs to bypass standard verification.
Management
Poor Platform Content Moderation
Social media platforms failed to proactively manage AI scams.
Regulatory Environment
Absence of Deepfake Regulations
Lack of strong legal frameworks to deter AI-generated scams.
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.
