Anthropic is killing off its Claude Cowork brand and product as a separate experience, instead folding all its capabilities back into the default Claude chatbot interface, removing the need for users to decide whether a request belongs in a conversational interface or a longer-running agentic workspace.
Beginning Wednesday, Claude will increasingly decide the distinction on its own, depending on the user's prompt and request. A user can start with a quick question, turn the same conversation into a multi-step project, and have Claude draw on the context, skills and connectors required to complete it.
“We built Cowork as a separate place for bigger work, and Design for visual work,” Anthropic says in its announcement. “People used both, and told us the frustrating part was deciding where a task belonged.” Now, that distinction now goes away.
The company is also launching a new collaborative document product called Claude Docs and a new presentation product called Claude Slides, both in beta, while bringing Claude Design directly into any chat session.
One caveat: Anthropic has used “Claude Docs” before, but not for this product. In September 2025, the company renamed its developer documentation site from “Anthropic Docs” to “Claude Docs” as part of a broader Claude brand consolidation. The document-authoring product announced Wednesday is new.
Anthropic did previously offer Claude for PowerPoint, which works inside Microsoft PowerPoint and received a major update in March 2026, but that is a separate product from the new native Claude Slides experience.
The changes begin rolling out to Claude Pro and Max subscribers on web, desktop and mobile over the coming weeks, according to Anthropic's announcement materials. Team and Free plans will follow. Users don't need to make any changes on their own.
The shift amounts to Anthropic retiring one of the most visible distinctions inside its flagship application: whether Claude is simply answering a user or being handed a job to execute.
Claude decides when a conversation becomes a project
Until now, Anthropic treated regular Claude conversations and Cowork as related but distinct modes. On web, desktop and mobile, users started both from the same Claude home screen but explicitly selected Cowork from a toggle button below the input box. In the Chrome side panel, opening the panel started a Cowork session directly.
Cowork was broadly available before Wednesday across Claude Desktop on paid plans, on the web and mobile for Pro, Max and Team customers and enabled Enterprise accounts, and through the Claude in Chrome side panel for Max and Team subscribers, with Pro access rolling out. Desktop offered additional capabilities because Claude could directly access local files and use the browser on the user's computer.
The new design removes that explicit Chat-versus-Cowork choice. According to background guidance from Anthropic, the Cowork toggle will disappear as the rollout reaches each account. When a conversation needs to turn into a longer-running task, Claude can do so without forcing the user to move their work or restart with a different mode.
Cowork's underlying behavior is not disappearing. Anthropic says users — including enterprises — who have already worked in Cowork retain their chats, projects, artifacts, connectors and skills.
Its agentic capabilities — including longer-running work, use of connectors and tools, task decomposition and the ability to keep working while the user is away — instead become part of how Claude operates generally.
Anthropic says the Cowork name itself will eventually disappear, although users may continue seeing it during the transition.
Claude Code, however, remains separate. In background guidance, Anthropic says Claude Code is intended for developers working from terminals and IDEs against codebases and is not part of Wednesday's consolidation.
Claude Docs and Slides turn outputs into working documents
The other major component of the release is a broader push from generating answers toward generating editable work products.
Claude Docs, launching in beta, lets a user ask Claude for a document and continue working on it collaboratively. Claude can draft sections, ask questions and comment on choices, while colleagues can edit the same document simultaneously. Documents begin private and can then be shared with named users or more broadly within an organization.
“Claude Docs” previously referred to Anthropic's developer documentation site after a 2025 rebrand, but obviously, the name was too enticing and descriptive to limit to those files, and will now apply to the Claude-created, shareable, collaborative web-based and downloadable documents the chatbot and underlying models can create.
Claude Slides follows a similar model for presentations. A user can request a deck in an ordinary Claude conversation, edit individual slides and present directly from Claude. This is also distinct from Claude for PowerPoint, Anthropic's existing PowerPoint add-in, which works within Microsoft's application rather than as a native Claude working document.
The key distinction is that Anthropic is treating both Docs and Slides as working documents inside Claude rather than simply files generated at the end of a chat. Anthropic says Docs, Slides and Design outputs can live at a single shareable link and remain editable, including from a phone.
Claude Slides and Claude Docs are web-native — they live at a shared link, like a Google Doc, rather than as a file on your computer.
As the company told VentureBeat in an emailed statement: "Claude Slides can be downloaded as PowerPoint or PDF, while Claude Docs can be exported to Microsoft Word or Google Docs, with additional export destinations planned."
Claude Design, previously its own product surface, now works from inside Claude conversations as well. Users can ask Claude for graphics, one-page visuals or other designs and then either manipulate elements themselves or tell Claude what to change.
That is an important operational distinction for enterprise users. The target is no longer simply “AI-generated content.” Anthropic increasingly wants Claude to own more of the workflow between a request and a usable deliverable.
The company illustrates that with a weekly reporting example: a user asks Claude to review pipeline movement, write a report in an established format, flag slippage and turn the highlights into five slides.
Claude can ask clarifying questions, continue working while the user is away and return both the document and presentation from the same underlying conversation, keeping the two outputs aligned. The task can then be scheduled to run again the following Monday.
Cowork started as “Claude Code for the rest of your work”
The merger comes only eight months after Anthropic introduced Cowork as a research preview on Jan. 12, 2026. At launch, it was much narrower: Cowork initially appeared as a Claude Max feature in the macOS desktop application, bringing the agentic pattern of Claude Code to nontechnical work.
Anthropic's idea was essentially to take the agentic architecture behind Claude Code and expose it to people who did not live in a terminal. The company had observed developers using Claude Code for decidedly non-developer work — vacation research, slide creation, email cleanup and other general-purpose tasks — and packaged the same delegation model behind a graphical interface. VentureBeat described Cowork at launch as extending Claude Code's agentic capabilities to nontechnical users.
Cowork's arrival also became an early catalyst in what traders and analysts later dubbed the “SaaSpocalypse,” although attributing the entire software selloff during summer 2026 to the January product launch would overstate the case. The initial Cowork release intensified an already developing debate over whether general-purpose AI agents could absorb work historically handled by dedicated software applications.
That concern grew sharper over the following weeks as Anthropic pushed Claude further into enterprise workflows and released more specialized capabilities. By late February, the broader market narrative had become difficult to separate from Anthropic's product cadence: VentureBeat reported that software companies including ServiceNow, Salesforce, Snowflake, Intuit and Thomson Reuters had come under pressure amid fears that increasingly capable AI agents could compress the value of standalone SaaS products.
Cowork helped start and crystallize the 2026 market debate over whether general-purpose AI agents could absorb work previously performed inside dedicated SaaS applications. The sharper “SaaSpocalypse” episode that followed, though, was not caused by Cowork alone. It reflected a broader accumulation of concerns around Anthropic's subsequent product releases, specialized plugins, Claude Code capabilities and the possibility that AI-native workflows could sit across — or replace parts of — existing enterprise software stacks.
That history makes today's product change more consequential than a navigation redesign. Anthropic is taking the agentic work model that originally lived behind a separate Cowork control and making it the default architecture of Claude itself.
Users can still control how much autonomy Claude gets
Anthropic is not removing user controls as it removes the mode boundary.
By default, Claude asks for permission before taking an action. Users can instead configure it to continue working and interrupt them only when closer review is required. Cowork already offered different approval modes for connector and tool use, and Anthropic's new unified product retains the basic idea that users determine how independently Claude should operate.
For organizations, those choices matter because Claude increasingly crosses application boundaries: reading files, using connected services, browsing websites, producing documents and presentations and scheduling recurring work.
Anthropic's bet is that users should no longer have to understand which internal product surface is appropriate before handing Claude that responsibility.
Under the new model, the interface becomes simpler even as the underlying system becomes more capable: start talking to Claude, and the product decides whether the job requires an answer, a document, a presentation, a design or a longer-running agentic process.
For Anthropic, that also represents a notable evolution from Cowork's January positioning. What began as “Claude Code for the rest of your work” is becoming, increasingly, simply Claude.