AIPhys.org Technologyabout 2 hours ago
Researchers Launch DivNet to Stop AI Attacks
3 min read
George Mason University professor Qiang Zeng and student Ying Meng developed DivNet, a method to defend voice assistants against mass attacks. Presented at the 2026 ACM MobiSys conference in Cambridge, England, the technique creates trillions of model variants to stop hackers. This approach reduces attack success while keeping recognition accurate and adding no processing delay.
Level
Hype check
Tap to vote and see what everyone thinks.
