VIDEO conferencing specialist Zoom Communications Inc is introducing new artificial intelligence (AI) management and security capabilities to give organisations greater control over how AI systems access business information and carry out workplace activities.
Unveiled at Zoomtopia 2026, the new capabilities include centralised AI management, sensitivity labels, AI guardrails and deepfake risk detection as Zoom expands the use of AI agents across collaboration, customer service and business operations.
The move reflects the growing importance of oversight and data governance as businesses increasingly rely on AI to automate tasks involving employees, customers and sensitive information.
Zoom founder and CEO Eric Yuan contended that understanding business context was essential for AI to deliver meaningful outcomes.
“AI models are only as powerful as their understanding of your business and that context makes all the difference,” he justified in a statement.
He went on to quantify that fragmented product experiences could isolate conversations and stall decisions.
Ensuring peace of mind
Zoom’s new Unified Context Layer is designed to address this by linking meeting discussions, chats, documents and information from connected applications, thus enabling AI agents to understand what needs to be done next.
Centralised AI Management will allow administrators to oversee how AI features are configured and used while sensitivity labels will help organisations enforce policies based on the classification of meetings and content, including through Microsoft Purview integration.
Elsewhere, AI Guardrails will enable account-level rules governing what AI can generate or use in responses while Deepfake Risk Detection aims to flag potentially synthetic media in meetings, hiring and customer interactions.
For Malaysian businesses adopting AI across customer service and internal operations, these controls could help address concerns over confidential information, compliance and responsible automation.
Moreover, Zoom is also introducing ZoomMate Agent Teams where specialised AI agents work together on tasks and My Notes Workflows which turn meeting summaries into follow-up activities in tools such as Zoom Chat and Salesforce.
ZoomMate Agent Teams is already available while sensitivity labels are available to selected Zoom Workplace customers. Centralised AI Management and Deepfake Risk Detection are expected in December 2026 with AI Guardrails scheduled for 1Q 2027. – Oct 22, 2026
