This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
A deepfake video featuring the voice and likeness of the late Martin Luther King Jr. endorsing Donald Trump was circulated on X by a pro-Trump account. The video was condemned by King's daughter, Bernice King, as 'vile' and 'fake.' The incident highlights ongoing concerns regarding the use of AI-generated synthetic media to influence democratic processes and voter perception during the 2024 presidential election.
A purported deepfake video allegedly depicting Martin Luther King Jr. endorsing Donald Trump circulated on X, where it was reported to have been viewed over 10 million times. The video, reportedly created by pro-Trump accounts, was reportedly condemned by King's daughter, Bernice King, who called it "vile" and demanded its removal.
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 a highly realistic but entirely fabricated political endorsement, spreading false information to millions of viewers.
The incident involves the coordinated use of AI-generated synthetic media to conduct a large-scale political influence campaign aimed at manipulating public opinion and voters.
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 intentionally generating and distributing a deceptive deepfake video using deployed AI tools to influence public opinion.
Added categories peak at 4: Severe (Epistemic Harm, Direct)
A high-fidelity AI deepfake of Martin Luther King Jr. endorsing a political candidate went viral days before the US presidential election. This incident highlights the substantial threat of AI-generated synthetic media in domestic information operations aimed at manipulating voter perception and eroding trust in democratic processes.
The AI system was used as a tool by human creators to generate a synthetic voice clone and deepfake video of Martin Luther King Jr. The reports describe the technology being used to produce specific outputs based on human direction rather than acting as an autonomous agent.
Limited Risk: The report describes an AI system used to generate deepfakes. Under the EU AI Act, providers of such systems must ensure that the outputs are marked as artificially generated.
The AI system was used to generate a synthetic video and a voice clone of Martin Luther King Jr. to make it appear as though he was delivering a speech endorsing Donald Trump.
The reports provide consistent, detailed accounts of the deepfake video, its content, the platform where it was shared, the number of views, and the reaction of the affected party (Bernice King). The role of the AI as a generative tool is clear, and the political context is well-documented.
None
While the specific AI generator model is unknown, this does not prevent confident classification of the incident's domain, causal factors, or harms.
Technology
Highly Convincing Voice Synthesis
AI voice cloning allowed realistic replication of King's delivery.
Unreliable Deepfake Detection
Detection tools have low accuracy, ranging from 25% to 82%.
Data Inputs
Historical Audio Training Data
Abundant public recordings of King enabled training of the AI voice model.
Human Factors
Intentional Deception by Creators
Political meme groups actively chose to create misleading AI content.
Emotional Vulnerability to Media
Audiences are susceptible to emotional manipulation by deepfakes.
Process and Methods
Ineffective Platform Moderation
Social media platforms rely on slow crowdsourced notes rather than removal.
Management
Prioritizing Engagement Over Safety
Platforms allow viral false content to spread to boost user engagement.
Regulatory Environment
Patchwork of State AI Laws
Lack of unified federal regulation creates gaps in penalizing 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.
