Event
Researchers from digital defence firm A Security have disclosed a critical vulnerability in Zoom that allows attackers to hijack devices through screen sharing. The flaw, located in the real-time annotation protocol, was identified using publicly available AI models in fewer than 20 prompts. Zoom has issued server and client-side patches for all supported operating systems, including Windows, macOS, Linux, iOS, and Android.
2 published stories · artificial-intelligence · source reliability 67%
Publicly available AI models identified the security gap in fewer than 20 prompts, prompting Zoom to release urgent patches across all supported operating systems.
Background
Researchers from digital defence firm A Security have disclosed a critical vulnerability in Zoom’s real-time annotation protocol, enabling attackers to silently hijack devices during screen-sharing sessions. The flaw was identified using publicly available AI models with fewer than 20 prompts, underscoring the rapid democratisation of hacking capabilities. Zoom has released server and client-side patches for all supported operating systems, including Windows, macOS, Linux, iOS, and Android. The discovery coincides with a broader industry trend where major AI platforms, such as Google’s Gemini and OpenAI’s ChatGPT, have surpassed one billion monthly active users, highlighting the increasing accessibility of AI tools for both legitimate and malicious purposes.
Researchers from digital defence firm A Security have disclosed a critical vulnerability in Zoom’s real-time annotation protocol, enabling attackers to silently hijack devices during screen-sharing sessions. The flaw was identified using publicly available AI models with fewer th
Fact cards
Key entities: Zoom, A Security, Omer Gull, Yossi Torati, WIRED, Ars Technica.
What is the nature of the Zoom vulnerability? It is a critical flaw in the real-time annotation protocol that allows attackers to hijack devices via screen sharing without victim interaction. How was the vulnerability discovered? It was identified using publicly available AI mode
Ars Technica | https://www.wired.com/story/a-zoom-screen-sharing-bug-let-anyone-take-over-other-devices-on-a-call | published 2026-08-12T13:37:12.000Z | primary source | WIRED | https://www.wired.com/story/a-zoom-screen-sharing-bug-let-anyone-take-over-other-devices-on-a-call | p
Early June: A Security researchers identify the critical vulnerability in Zoom’s real-time annotation protocol using publicly available AI models. 2026-08-12: A Security discloses the vulnerability and demonstrates the exploit; Zoom issues patches for all supported operating syst
Timeline
A Security researchers identify the critical vulnerability in Zoom’s real-time annotation protocol using publicly available AI models.
A Security discloses the vulnerability and demonstrates the exploit; Zoom issues patches for all supported operating systems.
FAQ
It is a critical flaw in the real-time annotation protocol that allows attackers to hijack devices via screen sharing without victim interaction.
It was identified using publicly available AI models in fewer than 20 prompts.
All supported operating systems, including Windows, macOS, Linux, iOS, and Android.
It highlights the democratisation of hacking, where tasks previously requiring large teams and months of work can now be achieved quickly with AI assistance.