Anthropic has introduced Claude Meetings, a workplace feature that allows the Claude assistant to join video calls as a visible participant and begin working on tasks while people are still discussing them. The company also announced Claude Whiteboards and Claude Sheets as part of a broader push into productivity software. During a meeting, Claude can follow the conversation, answer questions in the chat, draft a plan, create a timeline on a whiteboard and retrieve information from connected work applications. After the call, notes and action items remain available in claude.ai. The launch moves Claude from a tool that responds to prompts toward an assistant that can participate in collaborative work as it happens.
Claude becomes a visible participant rather than a passive recorder
The visible-participant design is significant because it makes the assistant’s presence explicit to people in the call. Rather than silently recording a meeting in the background, Claude is presented as a participant that can be asked to help with a task. Teams can describe a project and have the system begin organising the work while the conversation continues. For example, a discussion about building an internal application could produce an initial plan, a timeline and a list of unresolved decisions. The value is not simply transcription. It is the ability to convert spoken discussion into working documents that a team can review and refine before the meeting ends.
Claude Meetings can also use connected work applications to retrieve information that is relevant to the discussion. A team reviewing current performance may ask for last year’s figures, while a project group may need details from an earlier plan. This makes integrations and permissions central to the feature. The assistant should only retrieve information that the meeting participants are authorised to access, and it should make the source of important facts clear enough for people to verify them. If the connected application contains outdated or conflicting information, Claude may reproduce that uncertainty in its plan. Human review remains necessary before a draft becomes an approved decision or commitment.
Anthropic’s other workplace features, Whiteboards and Sheets, extend the same idea into different forms of collaboration. Whiteboards can help turn a discussion into a visual timeline or action plan, while Sheets gives users a way to work with structured information. Together, these tools suggest that Anthropic wants Claude to handle more of the work that normally takes place across separate office applications. The strategy puts the company in more direct competition with Microsoft and Google, which already provide integrated suites for meetings, documents, spreadsheets and collaboration. It also follows OpenAI’s recent workplace product announcements and Meta’s expansion of its Muse assistant.
Connected applications make permissions and source quality important
The move toward active meeting participation raises governance questions that are different from those associated with ordinary chat. Companies will need policies for when an AI assistant may join a call, whether all participants must consent, what information it can access and how long meeting materials are retained. Teams should also distinguish between a draft produced during discussion and an approved record of what the organisation decided. A model may misunderstand a speaker, combine separate suggestions or convert a tentative idea into a firm action item. Clear review workflows can reduce those risks, but the product should make it easy for participants to correct errors before outputs are shared.
Claude Meetings is initially positioned as a productivity feature rather than an autonomous decision-maker. Its usefulness will depend on whether it can reliably follow fast-moving discussions, preserve context across interruptions and create documents that require less editing than a manual summary. Enterprises will also assess security, data retention, administrative controls and integration support before allowing the assistant into sensitive calls. The broader competition is moving from who can produce the best answer in a chat window to who can help teams complete work across their existing systems. Anthropic’s launch is an attempt to make Claude a working participant in that process, but deployment quality will matter more than the demonstration.