This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video depicting Philippine President Ferdinand Marcos Jr. using drugs was released by political opponents to discredit him ahead of his State of the Nation Address. Authorities confirmed the video was fake, noting it was part of a broader pattern of AI-generated disinformation, including a previous audio deepfake, aimed at causing political instability. The incident has prompted calls for stricter regulation and prosecution of those using deepfake technology for political manipulation.
A purported deepfake video depicting Philippine President Ferdinand Marcos Jr. using drugs was released by Rodrigo Duterte ally Claire "Maharlika" Contreras at a Maisug rally in Los Angeles. The clip was reportedly timed to discredit Marcos ahead of his State of the Nation Address. Philippine authorities reportedly confirmed it was fake. The video was allegedly amplified by China's Spamouflage network via coordinated inauthentic accounts on X and YouTube.
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 false video and audio content depicting the President.
The deepfake video was deployed as part of a coordinated political disinformation campaign aimed at manipulating public opinion and destabilizing the government.
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 political opponents intentionally creating and distributing a deepfake video to destabilize the administration.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
Political opponents targeted Philippine President Marcos Jr. with a highly realistic deepfake video depicting drug use, timed to destabilize the government before a major national address. This incident highlights the growing threat of AI-enabled information warfare to sovereign government operations and democratic integrity, prompting immediate legislative and regulatory reactions.
The AI system was used as a generative tool to produce synthetic video and audio content based on user inputs, without exhibiting autonomous planning or decision-making capabilities.
Risk Level 3: Limited Risk. The report describes the use of AI-generated content (deepfakes), which falls under the Limited Risk category of the EU AI Act and is subject to transparency obligations.
The AI technology was used to generate a deepfake video of the President, and a previous incident involved generating a deepfake of his voice.
The reports provide clear, detailed, and consistent accounts of the deepfake incident, the political context, and the official government response. Forensic details proving the video was fake are explicitly described.
None
Technology
Deepfake Generation Technology
Advanced AI tools allowed creation of convincing fake video and audio.
Lagging Cyber Infrastructure
Philippine infrastructure is still catching up to handle advanced AI threats.
Data Inputs
Publicly Available Media Data
Abundant public video and audio of the President enabled AI model training.
Human Factors
Low Public Digital Literacy
Audience with little knowledge is easily persuaded by convincing deepfakes.
Malicious Political Motivation
Duterte supporters sought to destabilize the current administration.
Process and Methods
Weak Systemic Coordination
Government structures lack coordinated response to rapid AI disinformation.
Delayed Forensic Verification
Government had to rely on manual physical comparison to prove video was fake.
Management
Inadequate Proactive Risk Assessment
Government was unprepared for deepfakes targeting high-profile officials.
Regulatory Environment
Absence of Specific AI Regulation
Existing Cybercrime Act of 2012 lacks targeted rules for deepfakes.
Insufficient Election Guidelines
Election commission is only just drafting guidelines for AI use in 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.
