Anthropic launches Claude Meetings in beta, adding Claude to video calls as a working participant

Anthropic has launched Claude Meetings in beta, a feature that lets its Claude assistant join video calls as a visible participant that can follow the discussion, answer questions and produce work while the meeting is still running. According to Anthropic, the feature is rolling out on Team and Enterprise plans, with Google Meet supported first and additional video conferencing platforms planned.

A participant, not just a recorder

Claude appears in the participant list and posts a notice in the call's chat when it joins, Anthropic says. People on the call can ask it questions out loud or in writing; by default Claude replies in writing, and answers posted to the call chat are visible to everyone on the call. It can also reach into connected tools, with permission, to pull in information.

Claude can build artifacts live. In Anthropic's example scenario, a team planning a customer event asks Claude to start a document capturing the plan, then asks what last year's event cost. Claude requests approval to consult a connected tool, posts the figure in the meeting chat and compares it with the current plan. When the discussion turns to scheduling, Claude puts the timeline on a whiteboard that teammates can edit while the conversation continues.

Claude also handles notes and action items so no one on the call has to. Everything it produces, including notes, transcripts, documents and whiteboards, lands on a dedicated page for that meeting in claude.ai. Users can return to past meetings and ask Claude about them, or add Claude to the next one from the calendar in claude.ai. Claude joins when someone adds it or when auto-join is enabled.

Privacy guardrails

Anthropic has shipped the feature with conservative defaults. Claude stays out of calls when someone from outside the organization is invited, and leaves if an outside guest joins partway through, unless an admin allows outside guests. On calls with guests, Claude posts its notice in the chat but does not reply there, and what guests say is included in the notes and transcript.

Claude uses each person's connected tools only with that person's permission: it asks once and the approval carries forward. For spoken questions, Claude currently confirms who is speaking each time before it answers. Anthropic says Claude saves a transcript, a summary and screen descriptions, not audio, and advises organizers to make sure everyone on a call is comfortable being transcribed before adding Claude or turning on auto-join. Meeting pages are visible to invited people within the organization and can be shared internally.

Availability and cost

Google Meet support is available now, with other platforms to follow. On Team plans, inviting Claude is on by default and admins can switch it off; on Enterprise plans, an admin enables it for all or part of the organization. While in beta, the feature is unavailable to some organizations, including those with zero data retention, HIPAA-ready configuration or customer-managed encryption keys.

During the beta, notes and transcripts are free on Team and Enterprise plans, as are Claude's in-call answers, unless an answer uses one of the connected tools, which counts toward the plan's usage. Documents, whiteboards and other artifacts Claude creates in a call are free through February 10, 2027. Anthropic is also working with partners including Gong, Attention, Granola, Fireflies.ai, Otter and Plaud so meeting transcripts can flow into tools teams already use.

Why it matters

Meeting transcription is a crowded market, but Claude Meetings goes beyond passive capture: Claude is positioned as an active participant that acts, drafting documents, sketching whiteboards and looking up data mid-call. That pushes AI assistants deeper into live workplace workflows and raises the stakes on consent and access control, which Anthropic's defaults, such as avoiding external calls, skipping audio recording and requiring per-person tool permissions, appear designed to address. The beta carve-outs for HIPAA, zero data retention and customer-managed encryption keys show the compliance work still ahead, while free notes, transcripts and artifacts suggest Anthropic is prioritizing adoption over monetization for now. It is still a beta on two plan tiers and one platform, but the direction is clear: AI agents are moving onto the meeting invite list.