Deepfake bosses are crashing video calls, and…

In a sophisticated incident occurring in Hong Kong, an employee was targeted during a video conference where every other participant was a synthetic recreation of their colleagues. The attack leveraged stolen internal video material and AI-generated voice cloning to create a convincing, high-stakes environment that ultimately resulted in a significant financial loss. The victim only realized the deception after verifying the situation directly with their actual supervisor after the fact.

What happened

The attackers utilized advanced AI techniques to impersonate multiple familiar faces and voices, effectively creating a fake corporate meeting. By populating the call with several seemingly authentic colleagues, the criminals reinforced the illusion of legitimacy, making the request for a large financial transfer appear to be a standard business operation. The victim was successfully persuaded to transfer almost 24 million euros to the perpetrators.

This incident highlights a growing shift in cybercrime, where attackers are moving beyond simple face-swapping to orchestrate entire virtual meetings. Researchers at the Fraunhofer Institute for Secure Information Technology note that the presence of multiple familiar faces creates a powerful social engineering effect that overrides human judgment. While forensic research is currently focused on developing real-time warning systems to detect minute inconsistencies—such as facial blurring or signal irregularities—experts caution that these automated tools are not yet foolproof. The case serves as a stark reminder that even experienced professionals can be misled when trust is exploited through synchronized, AI-driven impersonation.