AI & Technology
On-Device Security: Scam.ai and Qualcomm Launch Real-Time Deepfake Detection for Enterprise Video
Unveiled at Computex 2026, the 'Halo' model processes video streams locally, offering high-stakes executives a powerful defense against synthetic identity fraud.
By 19Network Editorial Team · Jun 29, 2026 · 5 min read
As corporate identity fraud attempts skyrocket globally, a major new hardware-software partnership introduces edge-computing AI capable of identifying synthetic media instantly without compromising data privacy.
The ongoing war against malicious artificial intelligence entered a critical new phase today at the Computex 2026 exposition in Taipei. Cybersecurity firm Scam.ai , in direct collaboration with semiconductor leader Qualcomm, officially announced the commercial rollout of "Halo," a groundbreaking, on-device deepfake detection model designed specifically for live video communication. The launch addresses one of the most glaring vulnerabilities in modern corporate security: the manipulation of human identity via highly sophisticated, real-time synthetic audio and video. Enterprise security architectures have historically struggled to defend against deepfake infiltration. Once a human operator believes they are looking at their CEO, CFO, or a legitimate job candidate on a video call, traditional firewall and network security protocols become entirely obsolete. Industry data paints an alarming trend, showing that deepfake fraud attempts targeting corporate entities have increased by over 2,000% over the past three years. Compounding the issue, a recent human resources index revealed that only 31% of talent acquisition leaders feel properly equipped to detect identity fraud during…