· via dev.to (home feed)
Anthropic puts Claude into video calls with Meetings, Whiteboards and Sheets beta
Anthropic's Claude Meetings beta adds a visible AI participant to Google Meet calls that drafts plans, takes notes and builds artifacts in real time, alongside new Whiteboards and Sheets tools.

Anthropic moves Claude into meetings
On November 10, 2026, Anthropic announced a package of workplace features headlined by Claude Meetings, a beta that embeds Claude in video calls as a visible participant. According to a write-up on dev.to, the rollout began on November 12 for Team and Enterprise plan customers, and at launch the feature works only with Google Meet. Anthropic says support for other conferencing platforms is coming but has not named them or given dates.
The announcement also included Claude Whiteboards and Claude Sheets, and confirmed that Claude Data, previously called Claude Dashboards, is now generally available.
The beta carries restrictions for regulated customers: organisations using zero data retention, a HIPAA-ready configuration, or customer-managed encryption keys cannot use Claude Meetings yet.
An AI participant rather than an after-the-fact summary
Claude is added to a Google Meet call the way a colleague would be. As the conversation proceeds it drafts the plan being discussed while people are still discussing it, maps a timeline on a whiteboard as dates are agreed, and answers questions in the meeting chat, in writing by default so it does not interrupt anyone. It can pull real figures from connected work apps with the permission of the person asking, and it can build documents, whiteboards and other artifacts during the call that teammates refine together.
After the call, notes, a summary, action items, a transcript and every artifact Claude produced land on a dedicated meeting page on claude.ai.
The dev.to author frames the significance as merging two steps into one: the output of a meeting can be finished before the call ends, rather than written up days later.
Privacy defaults and guardrails
The dev.to report treats the privacy model as the centrepiece of the launch. Claude appears as a visible participant and posts a notice in the meeting chat when it joins. It saves a transcript, a summary and screen descriptions, not audio. By default it stays out of calls involving outsiders: it will not join when someone outside the organisation is invited unless an administrator explicitly authorises it, and it leaves automatically if an external guest joins partway through, including another company's notetaker bot.
Connected tools operate under each person's own permissions, so Claude can only reach data the person asking could already access. Anthropic is also explicit that transcription consent remains a human decision: participants should know and agree before Claude is added or auto-join is enabled.
Beta pricing
Notes, transcripts and in-call answers are free on Team and Enterprise plans during the beta, except that answers using a connected tool count toward plan usage. Documents, whiteboards and other artifacts created during calls are free through February 10, 2027. Anthropic has not said what the features will cost once they leave beta.
Whiteboards and Sheets
Claude Whiteboards, in beta on paid plans across web, desktop and mobile, turns a spoken plan into an organised visual workspace with timelines and deadlines while the conversation is still running, and can convert a finished board into a document or a slide presentation. Claude Sheets, also in beta on paid plans, lets teams create and edit spreadsheets collaboratively with Claude, removing the copy-and-paste step that sharing a Claude-built table previously required.
Why it matters
Claude Meetings is another step in AI assistants moving out of the chat box and into the software where work actually happens. The dev.to author points to a broader pattern, citing Meta's Muse placing calls to businesses and OpenAI giving ChatGPT a wallet, as evidence that agents are becoming participants in workflows rather than question-answering tools.
The immediate practical change is that a meeting's output, meaning the plan, the timeline and the action items, can exist by the time the call ends. That narrows the gap between what a team agrees verbally and what ever gets written down, which is where many projects drift.
The privacy defaults will likely decide adoption. Visible participation, no stored audio, automatic exit when external guests appear and permission-scoped tool access address the most common objections to meeting recorders. Still, organisations handling regulated data are excluded for now, and pricing after the beta is unknown. Teams evaluating the feature should begin with internal calls and inform participants in advance, as the source advises, since an in-chat notice alone does not constitute consent.
- #anthropic
- #claude
- #ai-agents
- #video-conferencing
- #productivity