Anthropic has put Claude inside the meeting itself. The AI assistant can now join a video call as a visible participant, follow the conversation, and start building documents and whiteboards while people are still talking. Claude Meetings, announced on 10 November and now rolling out in beta, is available to Team and Enterprise customers, starting with Google Meet. It takes notes, writes action items, answers questions asked out loud or in chat, and files everything on a meeting page afterwards. The launch lands in the middle of Anthropic’s push into workplace software, in the same week its public offering is drawing close attention.
What Claude does in the call
Say the team is planning a customer event, and Claude is on the call. Someone says, “Claude, start a doc for this plan.” Claude writes up what the team has said so far, and teammates on the invite can edit the plan while the discussion continues.
When the talk turns to budget, someone asks what last year’s event cost. Claude looks it up with approval, posts the figure in the meeting chat, and compares it with the current plan. When the discussion moves to timelines, someone asks Claude to put the schedule on a whiteboard, and the team works on it alongside the document.
Before the meeting ends, someone asks Claude to share a recap with the team. After the call, the notes, action items, plan and whiteboard sit on the meeting’s page in claude.ai, and anyone on the invite can go back later and ask Claude about a meeting they joined.
The guardrails
Anthropic has built the feature around consent and boundaries. Claude appears as a participant and posts a notice in the call’s chat, and it saves a transcript, a summary and screen descriptions — not audio. Before adding Claude or turning on auto-join, Anthropic says, teams should make sure everyone on their calls is comfortable being transcribed.
By default, Claude stays out of calls when someone from outside the organisation is invited, and it leaves if an outsider joins partway through. An admin can allow Claude on calls with outside guests, but on those calls it posts its notice without replying in the chat.
It uses each person’s connected tools only with that person’s permission. It asks the first time, and each time it is addressed out loud, it asks again to confirm who is speaking. An answer posted in the call’s chat is visible to everyone on the call.
Part of the office push
Meetings is the fourth piece of Anthropic’s workplace suite, after Claude Sheets, Whiteboards and Data, which launched on 10 November. Anthropic is also working with meeting-tool makers including Gong, Attention, Granola, Fireflies.ai, Otter and Plaud, so a meeting’s transcript can soon be sent to the tools a team already uses.
The beta is rolling out on Team and Enterprise plans only, with Google Meet the only platform supported at launch and others promised later. On Team plans the feature is on by default and an admin can switch it off; on Enterprise plans an admin turns it on.
Notes and transcripts are free during the beta, and the documents and whiteboards Claude makes in a call are free through 10 February 2027. The beta is not available to organisations with zero data retention, a HIPAA-ready configuration or customer-managed encryption keys.

Our Read
The meeting is the last place where AI has watched from the sidelines. Note-takers transcribe; they do not build the plan while the argument is happening. Anthropic is betting that the difference matters — that a model which can pull last year’s budget, draft the document and map the timeline in the same conversation is worth inviting into the room.
The comparison with xAI’s Grok Bot, which left beta for team meetings earlier this week, shows the race is already on. Google Meet already takes notes natively through Gemini, so Anthropic is competing on its rival’s home ground, and Microsoft’s Copilot sits in Teams calls too.
The honest test is trust, not capability. A meeting is where people speak loosely, disagree and change their minds, and a transcript that never forgets changes how freely they do that. Anthropic’s defaults — out of guest calls, no audio kept, admins in control — are the right shape.
But they put the burden on each team to agree the rules before the first call, and the beta’s limits narrow who can even try it: Google Meet only, paid plans only. The question is not whether Claude can do the work in a meeting, but whether teams will trust it to listen.
Why It Matters
If it works, the meeting stops being a dead end for decisions. The plan, the budget check and the action items already exist when the call ends, instead of being reconstructed from memory afterwards. That is a small change per meeting and a large one across a company that holds thousands of them.
It also moves Claude from a chat window into the flow of work, which is where the money in workplace software actually sits — and where Anthropic hopes to keep customers paying long after the novelty of a new model fades.
It matters for Anthropic too. The company is weeks from pricing its public offering, and its order book is already several times oversubscribed. A workplace suite that turns Claude into a colleague rather than a tool gives investors a story about recurring revenue beyond the model itself — provided teams actually let it into their calls.
Background
Anthropic announced Claude Meetings on 10 November alongside Sheets and Whiteboards, describing it then as coming soon. Its own article on 12 November confirms the beta is now rolling out. The launch follows the release of Claude Fable 5.2 earlier in the week, as Anthropic builds out its product line ahead of the offering.
Source: Anthropic’s announcement and TechCrunch’s report.


0 Comments