Israeli cybersecurity firm A Security said a researcher used publicly available AI models and fewer than 20 prompts to uncover Zoom vulnerabilities and build a working attack in less than 24 hours. The flaws affected Zoom’s annotation system and, according to the firm, could let an attacker control another meeting participant’s device without any action from the victim. A Security said the exploit was tested on Zoom apps for Windows, macOS, Linux, Android and iOS and tracked the bugs as CVE-2026-53413, CVE-2026-53414 and CVE-2026-53415. The finding carries particular significance for cryptocurrency users because hackers have repeatedly used Zoom and other video calls to target founders, investors and executives, often by impersonating trusted contacts and pushing fake updates or malware. A Security said Zoom released fixes between June 22 and July 20 after the first flaw was reported on June 10, but users still need to update because the company’s server-side safeguard could not filter malicious messages in end-to-end encrypted meetings.