Think Zoom calls are safe? AI tool flags dangerous security flaw that could expose your device
A vulnerability in Zoom’s screen-sharing feature could have allowed attackers to take control of devices without victim interaction, highlighting how AI is changing the cybersecurity threat landscape.

- Aug 12, 2026,
- Updated Aug 12, 2026 12:07 PM IST
If you regularly use Zoom for work, webinars, or personal conversations, then you may need to think twice before sharing your screen. Researchers have found that a lesser-known part of the screen-sharing feature could have created a serious security risk for you.
The flaw was discovered in early June by researchers at digital defence firm A Security while using publicly available AI models to analyse Zoom's software. According to the researchers, it took fewer than 20 prompts to uncover the vulnerabilities and create a working attack.
Must Read: Data breach costs for Indian firms hit record ₹25.5 crore as AI-driven cyber threats surge
How the Zoom vulnerability worked
The vulnerabilities were found in the protocol used for real-time annotations during screen sharing. Researchers said an attacker on the same Zoom call could potentially exploit the flaw to take control of another participant's device.
What made the vulnerability particularly concerning was that no interaction from the victim was required. Anyone on a call involving screen sharing, whether a participant or the host, could have been exposed to the silent attack.
Must Read: OpenAI makes GPT-5.6-cyber less restricted for trusted cybersecurity researchers: All details
The issue affected Zoom applications running on Windows, macOS, Linux, iOS and Android, according to the researchers.
For you, the risk went beyond losing control of your device. Researchers warned that an attacker who gained access to a company employee's computer could potentially obtain credentials and use them to move laterally through an organisation's network.
Zoom has patched the flaws
There is, however, an important reason not to panic. The vulnerabilities have already been patched, with Zoom rolling out fixes on both its servers and the applications installed on users' devices, according to WIRED.
The discovery also highlights a growing cybersecurity concern - AI tools are becoming capable of finding software vulnerabilities that previously could have required significant time and specialised expertise. Security researchers said the same discovery might previously have required a team of five people working for around six months.
For you as a Zoom user, the incident is a reminder that even familiar features such as screen sharing can become potential attack surfaces. Keeping your Zoom software updated is therefore important, particularly as security fixes continue to be released.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
If you regularly use Zoom for work, webinars, or personal conversations, then you may need to think twice before sharing your screen. Researchers have found that a lesser-known part of the screen-sharing feature could have created a serious security risk for you.
The flaw was discovered in early June by researchers at digital defence firm A Security while using publicly available AI models to analyse Zoom's software. According to the researchers, it took fewer than 20 prompts to uncover the vulnerabilities and create a working attack.
Must Read: Data breach costs for Indian firms hit record ₹25.5 crore as AI-driven cyber threats surge
How the Zoom vulnerability worked
The vulnerabilities were found in the protocol used for real-time annotations during screen sharing. Researchers said an attacker on the same Zoom call could potentially exploit the flaw to take control of another participant's device.
What made the vulnerability particularly concerning was that no interaction from the victim was required. Anyone on a call involving screen sharing, whether a participant or the host, could have been exposed to the silent attack.
Must Read: OpenAI makes GPT-5.6-cyber less restricted for trusted cybersecurity researchers: All details
The issue affected Zoom applications running on Windows, macOS, Linux, iOS and Android, according to the researchers.
For you, the risk went beyond losing control of your device. Researchers warned that an attacker who gained access to a company employee's computer could potentially obtain credentials and use them to move laterally through an organisation's network.
Zoom has patched the flaws
There is, however, an important reason not to panic. The vulnerabilities have already been patched, with Zoom rolling out fixes on both its servers and the applications installed on users' devices, according to WIRED.
The discovery also highlights a growing cybersecurity concern - AI tools are becoming capable of finding software vulnerabilities that previously could have required significant time and specialised expertise. Security researchers said the same discovery might previously have required a team of five people working for around six months.
For you as a Zoom user, the incident is a reminder that even familiar features such as screen sharing can become potential attack surfaces. Keeping your Zoom software updated is therefore important, particularly as security fixes continue to be released.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
