Claude Joins Your Google Meet Calls: Anthropic Launches Claude Meetings in Beta

Photo: Hippopx (public domain). Illustrative image.
What happened. Anthropic has put Claude inside the meeting itself. Claude Meetings, launched in beta on 12 November 2026, lets Claude join a Google Meet call as a visible participant: it answers questions out loud or in the chat, builds documents and whiteboards while the discussion is still going, and afterwards leaves notes and action items on a meeting page in claude.ai. It was teased as “coming soon” two days earlier, when Anthropic unveiled its office suite — Claude Sheets, Whiteboards and Data — and the beta is the first piece of that suite to actually sit in a live conversation.
What it does in the room. The pitch is that Claude is not a recorder bolted onto the call but a participant with the meeting’s context: it can be asked a question mid-discussion, pull an answer from the conversation so far, and turn the outcome into a document before anyone has written one up. After the call, the meeting page holds the notes and action items. What it keeps is deliberately narrow — transcripts, summaries and descriptions of shared screens, but no audio. And it has a hard boundary: by default it stays out of calls with outside guests, and leaves automatically if one joins.
The limits are the story for now. The beta runs only on Team and Enterprise plans — on by default for Team, admin opt-in for Enterprise — and only on Google Meet, with other platforms “coming soon”. Notes and transcripts are free during the beta, and the documents and whiteboards it creates are free through 10 February 2027. Organisations with zero-data-retention, HIPAA-ready or CMEK setups are excluded entirely. Transcript export goes to partners including Gong, Attention, Granola, Fireflies.ai, Otter and Plaud — which is also a quiet admission that Claude is entering a crowded room: Fireflies and ElevenLabs launched their own AI teammates on the same day this week, and the meeting is where the new “AI teammate” category is being fought.
Why it matters. A note-taker that only summarises is a convenience. An assistant that can answer, build and act during the meeting is a different product — and a different privacy question, since every participant is now talking in front of an AI that remembers the transcript. Anthropic’s answers so far — visible participant, no audio kept, auto-exit when outsiders join — are the right shape, but they are promises in a beta, not defaults a buyer can audit yet. The test is whether teams actually let Claude into the calls that matter, and whether the notes it leaves are good enough that nobody re-writes them.
Sources
- Anthropic — “Introducing Claude Meetings, now in beta” (12 November 2026): https://claude.com/resources/articles/claude-meetings
- FreshSynapse — “Claude Gets an Office Suite: Anthropic Launches Sheets, Whiteboards and Data”: https://freshsynapse.com/articles/anthropic-claude-office-suite-2026-11-11/
- FreshSynapse — “Fireflies Fred vs ElevenLabs ElevenAI: The AI Teammate Race Started on the Same Day”: https://freshsynapse.com/articles/fireflies-fred-vs-elevenlabs-elevenai-2026-11-15/
- FreshSynapse — “What Is an AI Teammate?”: https://freshsynapse.com/articles/what-is-an-ai-teammate/