AI deepfake scams reported in Hong Kong and elsewhere amid prevention warnings
An employee of Hong Kong-based Arup Engineering reportedly lost 200 million Hong Kong dollars in a scam involving AI-generated video calls impersonating company executives using deepfake technology. The incident prompted broader discussion of risks posed by deepfakes combined with data breaches and the challenges of preventing such fraud.
What happened
In 2024, an employee of Arup, an engineering company in Hong Kong, received a video conference invitation purporting to be from the company's UK chief financial officer, claiming a confidential transaction was required.
Multiple 'financial personnel' appeared in the conference with faces and voices closely resembling real people, but were produced using AI multi-person deepfake technology.
The employee transferred funds 15 times totalling 200 million Hong Kong dollars to five local accounts before querying the headquarters and discovering the fraud.
Reported impact
The incident was described as the most severe AI deepfake fraud case in Hong Kong's history in terms of losses incurred.
Index analysis
Deepfake technology combined with data breaches poses magnified dangers, as personal privacy in the hands of bad actors enables AI to create custom-tailored fraud schemes.
When deepfakes combine with data leaks to target individuals as precisely as advertising, society could face trust collapse and regulatory failure, with current prevention measures potentially unable to counter such scaled and personalized threats.
Governance controls
Countering deepfake harms requires multi-level, multi-stakeholder strategies spanning technology, law, education and social governance, including developing detection technology, strengthening legislation, and widely promoting prevention knowledge.
However, because AI encompasses multiple advanced technologies and public knowledge is stratified, large numbers of people unable to quickly master such technology remain inevitable victims of deepfakes.
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.