This is a research prototype. The data and analyses are preliminary and not yet validated — we'd welcome your .
Exiled Hong Kong pro-democracy activists in the UK and Australia have been targeted by a transnational harassment campaign involving AI-generated deepfake images. The materials, which include sexually explicit content and fake advertisements portraying activists as sex workers, were mailed to neighbors and associates. Victims described the campaign as psychological warfare intended to intimidate critics of the Chinese government, with police investigating the incidents as malicious communications.
Purported deepfake images of exiled Hong Kong pro-democracy activists were reportedly mailed to neighbors and others in the UK and Australia in late 2025. The letters allegedly depicted altered sexual images and false advertisements portraying activists, including Carmen Lau, as sex workers. Police described the materials as malicious communications. The incidents were described by victims and officials as an escalation of transnational harassment and intimidation.
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 neighbors and employers of the activists were involuntarily exposed to highly explicit, offensive, and toxic synthetic pornographic content.
The incident involves the targeted creation and distribution of humiliating, sexually explicit synthetic imagery of specific victims to intimidate and harass them.
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 harm was caused by human actors who intentionally used deployed AI deepfake tools to generate explicit images for the purpose of harassing and intimidating the activists.
Added categories peak at 2: Minor (Psychological Harm, Direct)
Exiled Hong Kong pro-democracy activists in the UK and Australia were targeted by a transnational harassment campaign using AI-generated explicit deepfakes. This incident represents a substantial national security concern, highlighting how foreign state-aligned actors leverage synthetic media for transnational repression, violating state sovereignty and undermining democratic protections on human rights.
The AI system functioned strictly as a tool, performing the specific task of generating synthetic images based on direct inputs provided by human users, without exhibiting autonomous planning or interactive assistant capabilities.
Limited Risk: The incident involves 'digitally altered images' and 'deepfaked images', which are explicitly categorized under Risk Level 3 as AI-generated content requiring transparency obligations.
The AI system was used specifically to superimpose the victims' faces onto explicit bodies to generate highly realistic deepfake images.
The reports from reputable news organizations provide clear, consistent, and detailed accounts of the harassment campaign, the specific nature of the deepfaked materials, and the psychological impact on the victims. The lack of a specific model name does not hinder the classification of the safety risks or harms.
None. The intent to harass and intimidate using synthetic media is unambiguous.
Technology
AI Deepfake Generation
AI tools used to superimpose target's face onto explicit bodies.
Realistic Image Synthesis
Advanced AI making fake explicit images look highly convincing.
Data Inputs
Publicly Available Photos
Using target's public photos as input for AI face-swapping.
Exposed Personal Data
Target's real address and details paired with the AI images.
Human Factors
Malicious Intent to Repress
Perpetrators weaponizing AI to intimidate political dissidents.
Targeting Female Activists
Exploiting gender-based harassment to maximize psychological harm.
Process and Methods
Anonymized Distribution
Mailing printed deepfakes from abroad to evade digital detection.
Inadequate Police Response
Law enforcement struggling to trace the foreign origin of AI media.
Management
Lack of AI Safety Guardrails
AI developers failing to prevent generation of non-consensual images.
State-Backed Campaigning
Coordinated orchestration of intimidation using digital tools.
Regulatory Environment
Weak Transnational Laws
Lack of global laws to prosecute cross-border AI-driven harassment.
Lax Generative AI Controls
Absence of strict legal mandates forcing AI tools to block NCII.
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.
