Zoom wants artificial intelligence to stop talking and start working: this is its new technological bet
The company has introduced a platform that connects meetings, emails, chats, and business apps so its AI agents can carry out tasks. It is also developing a tool to detect possible deepfakes and other new features.
Zoom wants to change how people use artificial intelligence at work. The company has announced a new set of tools that aims to do more than write down what people say in meetings or sum up conversations. Its goal is to let AI use information from different apps to organize tasks, start work processes, and help people finish work that still needs to be done.
The main new feature is called Unified Context Layer. Zoom introduced it at Zoomtopia, the company's yearly event. The platform brings together information from meetings, phone calls, chats, documents, and outside apps. This lets workers and AI agents find the information they need without having to piece together old conversations or search through different tools to find decisions.
This idea addresses a major problem with AI at work today. AI assistants can write text, sum up documents, and answer questions, but many tasks need information stored in several different systems. Zoom wants to bring this information together so its tools can follow conversations over time and turn them into real actions.
Our work revolves around conversations... and understanding the context is essential. 🚀
— Zoom (@Zoom) October 22, 2026
At Zoomtopia, our CMO Kimberly Storin explained why context is so important for getting work done. If you miss it, the conversation may have been pointless.
That's the idea behind Zoom's… pic.twitter.com/nryHI9yZeu
AI that connects meetings, chats, and apps
The new platform will support several features across Zoom's products. One is an improved version of My Notes, its note-taking tool. It will also work for face-to-face meetings through a separate mobile app and connections to compatible devices.
The company is also developing My Notes Workflows. This feature will let users set up actions to happen after meetings in tools such as Zoom Chat and Salesforce. Users will be able to explain in everyday language what they want to automate, without having to build every step themselves.
Another new feature is ZoomMate Agent Teams, which became available when the company made its announcement. These teams let users assign different parts of a task to several specialized AI agents, which can work at the same time. Zoom is also expanding AI-powered search across Meetings, Phone, and Chat, as well as more than 40 outside apps. These include ServiceNow, Workday, Google Drive, and Microsoft OneDrive.
Zoom wants these features to work beyond its own platform, too. Developers will be able to connect Zoom tools to other assistants and systems through programming interfaces and integrations using the MCP protocol. The company is planning integrations with Microsoft 365 Copilot, Claude, OpenAI Codex, and Meta Muse. Support for Google Gemini is planned for a later date.
Zoom is preparing a real-time deepfake detector
The company has also announced new safety measures as AI agents become more capable and gain access to business information. One planned feature will detect possible deepfakes. Expected in December 2026, it will check audio and video signals in real time to look for signs that content may have been changed or faked.
This tool will not provide a definite answer about whether a recording is real. Instead, it will warn users about possible manipulation. This is becoming more important in video calls, where someone could use fake audio or video to pretend to be another person and trick employees.