For users, the change may begin with a small question: should Claude answer, or should it take the job and keep working? Anthropic is removing that choice from the opening screen, folding its Cowork agent into the main Claude experience while adding native tools for creating documents and presentations.
From chatbot to workplace
The decision reflects a broader shift in how people use AI at work. A conversation that starts with a request for market research can become a connected-data analysis, a recurring report or a presentation for a meeting. Until now, Anthropic separated those experiences. Users seeking a conventional answer opened Claude, while those assigning a longer, tool-using project turned to Cowork.
According to Anthropic, Claude will now decide when a task requires connectors, planning, extended work or follow-up execution. Existing Cowork conversations, projects, artifacts, connectors and skills are expected to remain available. Claude Code, designed for terminal and integrated development environment work, will remain a separate product.
That simplification could make advanced capabilities easier to discover. It also removes a visible boundary between asking for information and authorizing action. The user may no longer need to understand which product can do what, but may have less clarity about when the system has moved from responding to operating.
Work objects, not just answers
Anthropic is also introducing Claude Docs and Claude Slides in beta. Docs is designed for collaborative drafting and editing, while Slides lets users create, revise and present decks within Claude. The results can be exported to familiar formats, including Microsoft Word or Google Docs for documents and PowerPoint or PDF for presentations.
The significance is less about file conversion than ownership of the workspace. Instead of producing text that someone must transfer into another application, Claude is positioning its output as an editable, shareable object that can continue to evolve.
The initial rollout targets Pro and Max subscribers across web, desktop and mobile. Team and Free access will follow, Anthropic says.
The unanswered question
This strategy places Claude closer to productivity suites whose strength has traditionally been the surrounding workflow: permissions, revision history, collaboration and predictable handoffs. AI can make those tools more fluid, but it can also make responsibility harder to see.
A human preparing a report usually knows when research ends and publication begins. An agent may gather information, consult connected systems and revise a document across several steps. That creates practical questions about approval, scheduling, data access and review.
Anthropic’s bet is that one intelligent interface will feel more natural than a collection of specialized products. Whether it earns trust will depend on how clearly Claude shows what it is doing, what it has changed and when a person must take control.
- Coolcaesar · CC BY 4.0
This article was written with the assistance of an AI system and published automatically.