Deepfake nude images used to harass exiled Hong Kong activists
AI-generated deepfake nude images of Hong Kong activists living abroad were created and distributed via anonymous letters and posters, targeting the activists and their family members. The materials contained fabricated accusations of sex work and were mailed and emailed to neighbors and employers in the United Kingdom and Australia. Police in both countries are investigating.
What happened
Liu Ka-man, a former Hong Kong district councillor subject to a red notice from Hong Kong Police's National Security Department, received anonymous letters at her previous address in Maidenhead, England containing five AI-generated deepfake nude images. The letters, postmarked from Macau, depicted Liu's face composited onto other women's bodies, including one showing simulated sexual activity, along with fabricated body measurements and the text 'Welcome to find me!'
Tso Chi-fung, a former Hong Kong legislator now in Australia, reported that his spouse received harassment targeting her as a non-public figure. A poster titled 'Lonely Hong Kong housewife' was created with their old photograph, listing fabricated sexual services and pricing in Australian dollars, and was emailed to Tso's employer and physically distributed to Adelaide residents. Australian police traced the threatening emails to an IP address in Hong Kong.
Reported impact
Liu Ka-man stated she felt frightened and shocked by the harassment, saying 'I am a woman and they are threatening me this way.' She characterized the incident as an escalation in cross-border repression that uses AI tools to target and destroy women's reputations.
Tso Chi-fung expressed anger that his spouse, a non-public figure, was subjected to targeted harassment for the first time. An Australian resident who received the poster reported feeling troubled and fearful when neighbors asked if his home had become a 'knock shop.'
Index analysis
The incidents involved the intentional creation of AI-generated synthetic sexual imagery depicting real individuals without consent, combined with defamatory false claims about their employment and contact information, distributed cross-border via postal mail and email to third parties including neighbors and employers.
The coordinated nature of the attacks on multiple exiled activists and family members, the use of forged postal addresses and fabricated details, and the apparent targeting of women for reputational destruction suggest organized harassment rather than isolated incidents.
Classification
Sources and evidence
Sources not yet published
This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.