Loading page…
Meeting participation, whiteboards and spreadsheets are in beta. Data dashboards and interactive partner apps are generally available, with app actions subject to user approval.
Listen to this story
Anthropic’s workplace release, reported by TechRadar on November 10, adds Claude Meetings, which can participate visibly in a video call and draft plans or other materials before the discussion ends. The rollout also brings shared Whiteboards and Sheets in beta, while Claude Data and Interactive Connectors are generally available. The shift is toward doing collaborative work inside Claude, not just generating meeting records: connected apps retain user permissions, and Claude actions require user approval.
Claude Whiteboards turns live discussions into shared plans, timelines and action items; Claude Sheets supports collaborative trackers and budgets with tabs, formulas and dropdowns.
Claude Data connects customer-management systems and data warehouses to dashboards, data stories and narrated videos; certified queries and live sharing remain in beta.
Interactive Connectors open supported apps—including Adobe, Atlassian, HubSpot, Salesforce, NetSuite and Workday—inside Claude.
Claude can join a video meeting as a visible participant and begin drafting work while people talk, rather than only deliver notes afterward. Anthropic’s new workplace release pairs that beta feature with shared whiteboards and spreadsheets, plus generally available data dashboards and interactive connections to partner apps, TechRadar reported on November 10.
Anthropic describes Claude Meetings as more than a transcription tool. It follows the conversation, takes notes, drafts plans and answers questions in the meeting chat. It can also outline timelines on whiteboards and use connected work apps while the call is still underway.
When the call ends, notes, action items and the materials Claude created are available in claude.ai. The distinction is timing: the company is offering to start the follow-up work during the discussion, not simply assemble a record for someone to act on later. Claude appears visibly in the call rather than operating as an unseen participant.
Engineers stopped asking AI for snippets and started handing it whole tasks, which freed them to spend their time deciding what to build. For most other teams, work has stayed much the same: the email gets written a little faster, the meeting comes with a summary.
Anthropic, in its announcement, quoted by TechRadar
Two companion tools give teams shared places to develop that work with Claude. Whiteboards integrates directly with Meetings, connecting the live conversation to a collaborative board. Sheets provides a spreadsheet environment for building and editing structured files together, rather than treating every output as a written response.
Meetings, Whiteboards and Sheets are all in beta. That separates the new collaboration tools from the other two parts of the announcement: Claude Data and Interactive Connectors are generally available. The release therefore combines experimental shared workspaces with data and app connections that have reached general availability.
Claude Data, previously called Claude Dashboards, connects to company information held in customer-management systems and data warehouses. It can turn those sources into dashboards, data stories and narrated videos. General availability does not cover every capability: certified queries and live sharing are among the features still labeled beta.
Interactive Connectors address a different part of the workflow. Full versions of partner applications can open directly inside Claude, where the assistant can take actions with user approval. TechRadar names Adobe, Atlassian, HubSpot, Salesforce, NetSuite and Workday among the supported apps. This puts the working application inside Claude, rather than limiting the connection to information retrieved from it.
Those connections respect existing user permissions. Administrators also get Claude auto-connect, usage controls and Smart Reports to monitor team productivity and costs. The permission boundary and approval requirement are distinct: opening an application inside Claude does not, under the announced design, give the assistant unrestricted authority to act in it.
Story updates
Anthropic has launched Claude Sheets in beta (announced November 10, 2026), embedding a shared spreadsheet workspace inside the Claude assistant for paid subscribers.
A request for a hiring budget can now become a spreadsheet that colleagues and Claude edit together, without leaving the AI app. Anthropic launched Claude Sheets in beta on November 10, 2026, for paid subscribers. It brings formulas, tabs and shared editing into Claude, but leaves advanced spreadsheet work to tools such as Excel.
Users can ask for a sheet in any chat or start from the Sheets tile in the Artifacts tab. Anthropic’s examples include campaign trackers, hiring budgets and forecasts built from deal data. Users can supply their own numbers rather than ask Claude to create a structure without source data.
The result is an artifact—a saved workspace inside Claude—where people and the assistant edit the same cells. Claude can add tabs, write formulas across them, apply conditional formatting and create dropdowns, checkboxes and charts. The beta includes 144 functions, while users can also type values and write formulas themselves.
Claude outlines each edit as it happens, and formulas remain in the cells for people to inspect. Users can select a range and ask a question about it, such as why a total increased. They can also request changes in plain language, including highlighting overdue items or adding a budget-remaining column.
Sheets begin private. Sharing lets colleagues work alongside the owner and Claude, with cell comments available for discussion. Version history can restore an earlier state, but that rollback also removes edits made after the restored version—not just a particular change by Claude.
Claude Sheets reverses the arrangement offered by Claude for Excel and Claude for Google Workspace. Those products bring the assistant to files in existing apps; Sheets puts the spreadsheet inside Claude. Anthropic recommends it for light, shared work, and points users back to those other tools when they need their full feature sets.
Existing data can come in through CSV imports, Excel or Google Sheets, including by pasting from those apps. Finished sheets can leave as XLSX or CSV files. A separate “Send to” option continues work in partner apps; Anthropic names Airtable, Asana, monday.com and Smartsheet among the destinations.
The launch limits are substantial. Each sheet can hold up to 100,000 cells, and Anthropic’s setup guide lists several gaps beyond that size ceiling:
Spreadsheet controls: No pivot tables, protected ranges, find and replace, multi-column sorting, filter views or multi-select dropdowns.
Excel-specific work: Macros, VBA, Power Query, DAX and print layout are unsupported. PDF export is also unavailable.
Live connections: Sheets cannot pull live data from connected apps. Documents, decks and dashboards also cannot use a sheet as their data source.
The beta is available on Pro, Max, Ultra, Team and Enterprise plans. Work in Sheets counts toward the subscriber’s existing Claude usage limits. Web and desktop support editing; mobile access is more limited. Although the launch announcement broadly lists mobile availability, the detailed guide specifies basic read-only access on phones.
Team organizations get Sheets switched on by default. Enterprise organizations do not: owners must enable it under Organization settings, then Artifacts, and Artifacts itself must also be enabled. Administrators can use custom roles to make Sheets available to only part of an organization, rather than opening it to everyone.
Anthropic on Nov. 10, 2026 introduced Claude Data, a new artifact type that turns connected business data into dashboards, narrated short videos and scrolling visual stories. It builds SQL queries, exposes them for inspection, and saves creations to an Artifacts tab.
A question about company revenue can become more than a chart in Claude. Anthropic introduced Claude Data on November 10, 2026, as an artifact type for connected business data. The company says it can produce dashboards, visual stories and narrated short videos, with beta tools for reusing approved queries and sharing results across a team.
This is not a clean-sheet replacement for Claude’s existing dashboard tool. Anthropic’s help guide says Claude Data was previously called Claude Dashboards, and dashboards made under that name keep working. The introduction puts those dashboards alongside other ways to explain the same underlying numbers.
Anthropic is rolling out embedded partner apps inside Claude so connected tools can open as full workspaces alongside (or instead of) chat, and Claude can make edits users approve in-place.
A connected app can now become the workspace inside Claude, rather than a small preview in a conversation. Anthropic began rolling out embedded partner apps on November 10, letting users open supported tools beside chat or full screen and ask Claude to change the page, row or file they are working on.
Earlier this year, Anthropic introduced interactive connectors that brought small, live views of outside tools into Claude. This update changes how much of the tool is available: where a connector supports it, users can open the app itself, keep the conversation alongside it, hide chat or switch to a full-screen view.
Loading discussion...
Users connect a data warehouse, such as Snowflake or BigQuery, or an app already connected to Claude, such as Salesforce. They can then describe a question in a chat or open Data from the Artifacts tab. Claude writes and runs SQL—the language used to retrieve database information—for each chart.
Each chart exposes its query and the time its data was last refreshed. Users can inspect what Claude counted and ask for a correction if it does not match their intent. The guide describes several output formats beyond dashboards:
Scrollyteller builds a scrolling story from the numbers, rather than a single dashboard view.
Visual short turns the data into a short narrated video that walks through the figures.
Visual analysis provides a written summary of the numbers. Creations are saved in the Artifacts tab.
Refinement can happen through conversation or direct edits to titles, captions and charts. Version history includes those direct edits, and restoring an earlier dashboard changes it for everyone who opens it. View-only users cannot edit or restore versions.
Anthropic positions dashboards for quick exploration, not deeper analysis, which it directs users to continue in another tool. Refreshing also has a specific limit: data is refreshed when someone opens a dashboard and it is due for an update, rather than on a fixed schedule.
Certified queries, now in beta, let Claude reuse queries approved by a data team. Charts using them carry a badge. Teams can certify queries inside Claude or use definitions from dbt, Looker or Cube through existing connectors. Editing a certified query removes its certification; removing certification also strips the badge from dashboards that use it.
Certification is not restricted to a named data-team role: anyone with write access to a Data artifact can certify a query. Organizations keeping certified queries in a Git-backed plugin instead manage them through GitHub pull-request review.
The live-sharing beta is limited to the exact queries a dashboard needs, which cannot change without an editor’s approval. Anthropic tells users to check what a dashboard shows before sharing it this way. Creations start private; the guide also allows artifact sharing outside an organization, while describing live sharing specifically for organizational colleagues.
Claude Data is available on Pro, Max, Ultra, Team and Enterprise plans and counts toward existing usage limits. On Team, the feature and both betas—certified queries and live sharing—are on by default.
For Enterprise, Claude Data is scheduled to turn on by default on November 19, 2026, unless an owner has already saved a setting. Both beta features remain off by default. Owners manage the controls in organization settings, and Enterprise owners can limit Claude Data to specific groups using custom roles.
For Confluence and Notion, Anthropic says that means the full editor, signed into the user’s own account, rather than a limited preview. Users can also open a connected tool without first writing a prompt. Claude’s help is available from that same screen when they want it.
Claude can make requested changes directly in an open page, row or file. Users approve those actions through the connector’s existing approval process, and changes appear in the tool as Claude makes them. The user can keep working alongside it, rather than wait for a separate response to copy back.
The starting group includes Asana, Canva, Carta, Dropbox, monday.com, Smartsheet, Vercel and Webflow. Anthropic also describes several editing and business workflows available through other supported tools:
Confluence: Open a page beside chat, ask Claude to write or edit, and review changes live on the page.
Box: Ask Claude to revise an open file, then compare versions side by side before keeping the changes.
Adobe: Request photo retouching, edits and effects, or highlight, edit and redact PDFs.
Notion and HubSpot: Create a Notion document or database, or prioritize pipeline tasks and follow-ups in HubSpot and act on them.
Users connect these tools through Claude’s directory, then open them from chat or pin frequently used apps to the sidebar. Pinned tools can sit alongside artifacts and are also accessible through the + menu.
The rollout covers web and the Claude desktop app across all plans, but availability is not immediate for every account. It arrives with the new Claude experience that combines chat and Cowork, once an account is updated or an organization enables it.
For organizations that already allow all of a connector’s tools, the new features appear as they roll out. Administrators can continue restricting a connector’s tools using the controls they already have.
The underlying technology is MCP Apps, an open extension to the Model Context Protocol. It lets a connected service supply its own interface inside a compatible AI application. Anthropic says an app built this way is not limited to Claude: it can run in other AI applications that support the extension.
Claude now supports more of that extension, including app tools and embedded sites, so developers can bring fuller applications into the interface. Support for these expanded capabilities is in public preview, and developers can begin building today. Figma, Harvey and Shopify are also building integrations; Anthropic names them as work underway, not tools already available in this rollout.
Join the conversation
Explain when getting a head start would help—or create rework.
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.