This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Spain's People's Party (PP) circulated an AI-generated video on official social media channels that used deepfake technology to impersonate political rivals, including the prime minister's wife and other Socialist party figures. The video attributed false, defamatory statements to these individuals, prompting a legal complaint from a former party member who reported suffering from anxiety and social isolation as a result. While the party defended the content as political satire, the incident has sparked significant debate regarding the ethics and legality of using AI-generated disinformation in political campaigns.
Spain's conservative People's Party reportedly circulated an AI-generated video impersonating the prime minister's wife and Socialist figures making apparent admissions of misconduct. Former Socialist party member Leire Díez filed a defamation complaint; her lawyers alleged insults, anxiety requiring treatment and isolation at home with her daughters. The party reportedly defended the video as political criticism. Its AI disclosure was reportedly small and faint.
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 falsely depicted individuals admitting to illegal acts like soliciting prostitutes and hiring thugs, exposing the public to toxic/defamatory content and leading to harassment of the victims.
The video was a political campaign ad distributed on official party channels to influence public opinion and the upcoming general election.
The incident involved the unauthorized use of public likenesses to impersonate and target specific political figures with defamatory statements for political gain, fitting 4.3.
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 humans (the People's Party) intentionally using deployed AI deepfake tools to create and distribute a political campaign video targeting rivals.
Added categories peak at 2: Minor (Epistemic Harm, Direct)
Spain's conservative Peoples Party utilized AI deepfake technology to generate a defamatory campaign video targeting political rivals, including the Prime Minister's wife. While domestic in nature, the incident marks a significant escalation in the use of synthetic media for political disinformation, threatening information security and electoral integrity.
The report describes the AI system being used by the People's Party to generate a deepfake video. The AI performed a specific, well-defined function (generating synthetic video and voice) based on direct human inputs, operating as a tool without autonomous planning or agency.
Risk Level 3: Limited Risk. The report describes 'AI-generated content such as deepfakes' which are subject to transparency obligations. The PP attempted to meet this by noting 'at the bottom – in tiny, nearly transparent font – that it had been created with AI'.
The AI system was used to generate a synthetic video manipulating the likenesses and voices of political figures, which is the definition of deepfake video and voice generation.
The reports provide consistent, detailed accounts of the incident, including the specific video content, the platform views, the legal complaint filed by Leire Díez, and the political context in Spain. There is no significant ambiguity regarding what occurred or the harms caused.
The PP argued the video was political satire rather than malicious disinformation, though this does not change the factual occurrence of deepfake generation.
Technology
AI Deepfake Generation
AI was used to manipulate likenesses and voices of political figures.
Imperceptible Disclaimers
The AI disclaimer was written in tiny, nearly transparent font.
Data Inputs
Use of Personal Likenesses
Real photos, videos, and voices of politicians were used as inputs.
Human Factors
Intentional Disinformation
Creators intentionally put false admissions into the mouths of rivals.
User Confusion
People not used to AI can easily confuse deepfake videos with reality.
Process and Methods
Inadequate Content Filtering
Social media platforms allowed rapid spread of unverified deepfakes.
Political Campaign Satire
Using deepfakes as a campaign strategy to dominate online conversation.
Management
Prioritizing Virality Over Ethics
Party management approved deepfakes to counter rival social media reach.
Lack of Internal AI Guidelines
No ethical review process within the political party before deploying AI.
Regulatory Environment
Satire Regulation Loophole
AI regulations contain exceptions for satire, complicating enforcement.
Lack of Clear Disclosures
Weak enforcement of readable disclosure standards for AI-generated 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.
