Tech‘Zoomsday’ hack uncovered with fewer than 20 AI prompts

Zoom has fixed a major security flaw that could let an attacker take control of any participant’s device during a meeting. Researchers at A Security said they found the vulnerability by using fewer than 20 prompts with publicly available AI models. The exploit relied on Zoom’s annotation feature. The discovery was previously reported by Wired.

Zoomcybersecurityartificial intelligencesecurity vulnerability

TThe Verge★★★☆☆2026-08-11 14:45Original