Zoom launches AI tools to link chats, meetings & work

Zoom has introduced a Unified Context Layer and a new set of AI products for its workplace and customer platforms as part of a broader push to connect conversations with follow-up work across its software.

The new layer is designed to connect information from meetings, chats, documents and third-party applications, giving staff and AI systems a shared view of activity. Zoom says this should reduce the time teams spend reconstructing discussions, decisions and next steps across separate tools.

New meeting features include personalised transcripts and summaries for missed sessions, searchable notes linked to slides and chats, and workflow tools that turn meeting summaries into actions in applications such as Zoom Chat and Salesforce. Zoom also unveiled a refreshed Zoom Chat experience and ZoomMate Agent Teams, described as groups of specialised AI agents working together on tasks.

Zoom is also extending access to its tools beyond its own application. Through APIs and Model Context Protocol servers, developers and AI agents will be able to use functions including scheduling, transcription, search, administration and contact centre actions in other software environments.

This includes links to AI assistants such as Claude, Codex, Copilot and Muse, as well as search and query tools connected to the Unified Context Layer. The aim is to let businesses use Zoom data and functions within existing workflows rather than forcing staff to switch constantly between systems.

Alongside those changes, Zoom announced a new set of meeting and communication features, including voice translation using a synthesised version of a speaker's own voice, video pose correction, accent smoothing, and a live meeting AI agent intended to help participants find answers and keep discussions on track.

Customer systems

In customer service, Zoom is adding customer profiles to Zoom CX to provide real-time context across channels and interactions. It is also releasing transcript analysis tools for voice, chat and email, as well as a kiosk version of Zoom Virtual Agent for use in physical retail settings.

Zoom said customer context is often lost as interactions move between live agents, automated systems and different channels. The additions are intended to preserve a fuller history of previous contacts so customers do not have to repeat information and agents can see a broader record of past exchanges.

Zoom also outlined a broader set of administrative and risk controls tied to its AI products. These include centralised AI management, sensitivity labels, account-level rules for AI responses, deepfake risk detection and compliance support for AI-generated notes and summaries.

The controls are intended to help organisations manage access to information and align AI use with internal data policies. Some features are available immediately, while others are scheduled for release over the coming months and into next year.

Revenue target

Zoom expects to surpass USD $5 billion in annual revenue this year. It also said the number of Zoom Virtual Agent customers has risen by 256%.

The company cited user research indicating that customers using its AI tools report completing tasks twice as fast, with 72% saying the business impact was high or extremely high, compared with a market average of 63%. It added that adoption has broadened across sectors including healthcare, financial services, retail and education.

Eric Yuan, Founder and Chief Executive Officer of Zoom, linked the product changes to the company's broader AI strategy. "AI models are only as powerful as their understanding of your business, and that context makes all the difference," Yuan said. "Fragmented product experiences isolate conversations, stall decisions, and drag businesses down. Today, with the Unified Context Layer, we're laying the groundwork for a system of action designed to turn conversations into completed work."

Oracle Cloud Infrastructure was among the customers Zoom cited during the launch. Christine Sarros, Senior Vice President of Enterprise Engineering at Oracle Cloud Infrastructure, said teams often have the right information but struggle because it is spread across systems and notes.

"Our teams have the information they need, but it's often dispersed across meetings, notes, and different systems," Sarros said. "Zoom's new solutions could help bring that context together, so teams can enter customer conversations with a clearer understanding of what's been discussed, what's been decided, and what needs to happen next. We see an opportunity to spend less time reconstructing history and more time helping our customers move forward."