Anthropic's Claude Meetings Joins Google Meet Calls to Build Working Documents Live
Anthropic puts Claude inside Google Meet calls as a visible participant that drafts docs, timelines, and action items while you talk.
- Anthropic launched Claude Meetings in beta for Team and Enterprise plans on Google Meet
- Claude joins as a visible participant, drafting docs, timelines, and whiteboards live during calls
- Post-call page collects transcript, notes, action items, and any artifacts Claude built
- Safeguards include chat self-identification and auto-leaving if unapproved external guests join
- Shipped with Claude Whiteboards, Sheets, and Data, plus connectors for Adobe, Atlassian, and HubSpot
- Positions Anthropic directly against Microsoft 365 Copilot and Google Workspace in office productivity
Anthropic has announced Claude Meetings, an assistant designed to join Google Meet calls as a visible participant. It can answer questions, draft plans, build timelines, and create other working documents while people talk. Initial access is tied to Team and Enterprise plans, though Anthropic’s materials leave the boundary between the current beta and forthcoming features unclear.
The product expands the familiar meeting-assistant workflow. Along with producing a transcript and summary, Claude can create editable work during the call. Participants can inspect and correct that output before the meeting ends, reducing the follow-up required to turn a discussion into a project plan, document, or task list.
Claude works in the room
Teams invite Claude to a Google Meet call as they would another participant. It appears as a tile, identifies itself in the meeting chat, and accepts spoken or typed requests. A side panel displays its work so attendees can review and revise drafts in real time.
- Draft documents, timelines, and whiteboards from the discussion.
- Answer questions by voice or chat using connected business applications.
- Retrieve current figures from services Claude has permission to access.
- Record decisions, action items, owners, and unresolved questions.
- Store the transcript and generated materials on a dedicated page in Claude.
Nicholas Lin, Anthropic’s product lead for knowledge work, told Axios that the company tested Claude Meetings internally for about two months and used it in nearly all of its meetings. Anthropic has not published accuracy, latency, or task-completion benchmarks from that trial.
Sensitive calls get attendance controls
Claude announces its presence in the meeting chat and will not join a call containing external guests unless an administrator has authorized that use. If an unapproved external participant enters later, Claude leaves automatically. These defaults limit accidental exposure during client calls and other mixed-attendance meetings.
Organizations still need policies for participant consent, transcript retention, generated documents, and connected application data. Administrators also need to verify which users can invite Claude, which services it can query, and whether generated materials inherit the access controls of their source systems.
The beta label needs parsing
Anthropic’s announcement and launch coverage describe the rollout differently. The company’s page labels Claude Meetings as “coming soon,” while help material and reports from TechCrunch, Engadget, and Unite.AI describe beta access for parts of the Google Meet experience.
| Capability | Reported status | Availability |
|---|---|---|
| Live whiteboard in Google Meet | Described as a beta | Team and Enterprise plans |
| Full Claude Meetings participant | Anthropic says “coming soon”; some coverage describes a beta | Exact rollout scope remains unclear |
| Zoom and Microsoft Teams support | No release announced | Unavailable in the stated launch scope |
Administrators should confirm which controls and capabilities appear in their Claude workspace before planning a deployment. Google Meet is the only conferencing platform named for the initial release, and Anthropic has not provided a schedule for Zoom or Microsoft Teams.
Anthropic widens its office-suite push
Claude Meetings arrived alongside Claude Whiteboards for collaborative brainstorming, Claude Sheets for building spreadsheets and financial models, and Claude Data, the renamed dashboard product. Anthropic also added connectors for services including Adobe, Atlassian, and HubSpot.
The combined release places Claude deeper inside everyday enterprise workflows. Google Meet already provides Gemini-powered note-taking, Microsoft distributes Copilot across Microsoft 365, and services such as Otter, Granola, and Fireflies handle meeting capture. Claude’s distinguishing feature is live production of editable plans, tables, timelines, and other work products within the conversation.
That workflow can remove a common handoff: someone no longer has to copy a transcript into a document, spreadsheet, or ticketing system after the call. A manager or technical lead can review the draft while the relevant participants are present and assign corrections immediately.
What teams need to test
Before enabling Claude Meetings broadly, administrators and engineering leads need to test its behavior against existing security, compliance, and workflow requirements.
- Access and consent: Define who can invite Claude, how participants are notified, and which calls require explicit approval.
- Retention: Align transcripts and generated documents with meeting-recording, deletion, legal-hold, and data-residency policies.
- Connector permissions: Limit Claude to the applications and records required for each workflow.
- Output quality: Test technical jargon, overlapping speakers, acronyms, numerical claims, and task ownership. Assign a person to approve generated work.
- Interoperability: Verify export formats and downstream automation. The announcement does not specify a developer API, webhook support, or integrations with Zoom and Microsoft Teams.
For Google Meet organizations already using Claude, the practical gain is a shorter path from discussion to an editable plan. Adoption will depend on the rollout’s actual scope, the reliability of live output, and controls that fit each organization’s data policies.