Artificial intelligence assistants are increasingly being asked to do much more than answer questions. People now use them to research topics, organize projects, analyze information, prepare reports, write proposals and plan presentations, but the final stage of that work has often happened somewhere else. A response might begin in an AI chat, move into a word processor for editing, then into presentation software for slides before being shared with colleagues. Anthropic’s latest changes to Claude are designed to reduce those transitions by turning the assistant into a more complete workspace where conversations, documents, presentations and longer-running tasks can remain connected.
The company is merging Claude’s regular chat experience with Claude Cowork while introducing Claude Docs and Claude Slides directly into the same environment. Claude Design is also becoming available from normal conversations, giving users several ways to turn an idea into something more finished without first deciding which separate Claude tool they need to open. Anthropic says the goal is simple: users should be able to describe what they want done, while Claude determines which capabilities are appropriate as the task develops.
The change is significant because it pushes Claude beyond the familiar chatbot model. A conversation can now begin with a quick question, grow into research, become a written document and eventually produce a presentation based on the same information. Instead of treating each output as a separate project, Anthropic is trying to keep the context and work together from beginning to end.
Chat and Cowork are becoming one Claude
Anthropic previously separated everyday conversations in Claude from more involved work handled through Cowork. That made sense when the two experiences were aimed at noticeably different types of tasks, but it also created another decision for users: should a request begin as a normal chat or as a larger Cowork project?
The new approach removes much of that distinction. Anthropic says users can begin with a simple request and allow Claude to take on more of the work if the task grows in complexity. Claude can ask follow-up questions when information is missing, use available connectors and skills, continue working on longer assignments and present the resulting work inside the same broader experience. The company also says tasks can continue after a user steps away from a computer, with progress available again from another supported device.
That could make Claude easier to use for people who are interested in AI but do not want to think about product modes. A short question still behaves like a short question. A request such as preparing a weekly business report, however, can develop into a longer workflow with research, a document and presentation slides without requiring the user to start over in another part of the product.
Anthropic describes the new experience as one in which Claude can decide what a task needs. That is a subtle but important change in how AI productivity tools are being designed. Instead of asking users to learn a collection of separate AI features, the interface increasingly tries to hide that complexity behind the conversation itself.
Claude Docs brings writing and editing into the conversation
Claude Docs is one of the clearest examples of that direction. Rather than asking Claude to draft a report and then copying the result into a separate word processor, users can create an editable document directly from the conversation and continue working on it there.
A document can begin from a prompt, existing information or the context already established in a conversation. Claude can help structure the material, write sections, revise wording and respond to comments while the user retains the ability to edit the document directly. This creates a workflow that feels closer to working inside a collaborative editor than repeatedly exchanging large blocks of text with a chatbot.
Anthropic’s documentation says users can comment on content and mention Claude when they want a specific revision. Claude can respond to the comment and make changes to the relevant material, while people can continue editing the document themselves. Multiple collaborators can also work around the same document, making Claude another participant in an existing editing process rather than a separate destination for generating text.
Docs is also intended for more than plain text. Anthropic says documents can include visual elements such as charts, graphs, diagrams and timelines when appropriate. If Claude has access to relevant information through the conversation or supported connected sources, that material can become part of the document instead of being recreated separately.
When the work is finished, users are not required to keep it inside Claude. Documents can be downloaded or moved into familiar productivity formats, allowing Claude to serve as the place where the work is developed without necessarily becoming the only place where it can be used.
Claude Slides turns the same information into a presentation
Presentations are getting a similar treatment through Claude Slides. Users can ask Claude to prepare a deck from notes, research, an existing document or material that has already been discussed in the conversation, then continue refining the presentation without starting again from a blank slide.
That connection between Docs and Slides could be one of the more practical parts of the new system. A user might spend time building a detailed report in Claude Docs and then ask for a shorter presentation aimed at a meeting. Because both outputs can originate from the same conversation and source material, the presentation can reflect the conclusions and structure of the report rather than requiring the user to manually summarize everything a second time.
Anthropic says presentations can be edited directly, presented from Claude and downloaded as PowerPoint or PDF files. That makes Slides relevant even for teams that continue using traditional presentation software, because Claude can handle more of the initial creation and revision process while PowerPoint remains available for final distribution or additional editing.
AI assistants have been able to suggest slide outlines for years, but generating an outline is different from producing the actual presentation. The newer approach moves closer to the second option. The user can focus on what the presentation needs to communicate, while Claude takes on more of the work involved in turning that information into a structured deck.
Claude Design adds a visual layer to the same workspace
Documents and slides are not the only output formats being brought closer to the main Claude experience. Claude Design can now work inside conversations as well, extending the same workflow toward visual material.
Anthropic groups Docs, Slides and Design around the broader concept of work that can be created, edited and shared from Claude. This means a single conversation can potentially result in a written report, a presentation and supporting visual material without requiring each item to begin as an unrelated task.
That may be particularly useful for projects that naturally involve several formats. A small business preparing a new proposal, for example, might first organize research and notes, turn them into a document, create a short presentation for a meeting and then prepare a visual summary. Keeping those outputs connected to the same conversation reduces the need to repeatedly explain the project from the beginning.
It also suggests that Anthropic sees Claude’s future as something broader than a text interface. The conversation remains the starting point, but the finished result does not necessarily have to look like a conversation at all.
Connected services can bring existing information into the workflow
A workspace becomes considerably more useful when it can work with information that already exists elsewhere. Claude supports connectors and integrations that can make outside material available when a task requires it, allowing users to build documents and other outputs around existing files, business information and project context instead of relying entirely on information pasted into the prompt.
Anthropic gives examples in which Claude can retrieve information, work with existing resources and use that material while preparing larger assignments. Because Chat and Cowork are becoming part of the same Claude experience, those connectors, skills and existing context can follow the task as it develops rather than being limited to a separate mode.
Imagine preparing a weekly project update. Claude could work from information that is already available through connected tools, organize the important changes, prepare a written summary and then create several slides for a meeting. The value comes less from any one individual feature and more from reducing the number of manual transitions between collecting information and producing the final output.
That is also why Docs and Slides matter more when viewed as part of the wider Claude system. A document generator on its own is useful. A document generator connected to research, files, previous conversations and presentation tools becomes a more complete productivity environment.
Longer tasks no longer have to feel like extended chat sessions
Another part of Anthropic’s new approach is the ability for Claude to handle work that takes longer than a normal response. The company says users can hand over a larger task and return later, rather than keeping the conversation open while every step is completed.
Anthropic’s launch example describes a recurring weekly report that Claude can prepare, turn into a document and summarize into several presentation slides. The work can continue while the user is away, and the resulting materials can be checked later from another device. Tasks can also be scheduled to run again, which moves Claude closer to handling repeatable workflows rather than only responding to one request at a time.
That could change how people think about AI assistants. Instead of opening Claude only when they need an answer, users could increasingly hand it recurring pieces of work that have a clear structure. A regular report, project summary or presentation update becomes something Claude can help prepare repeatedly rather than something recreated manually every week.
The user still controls the final result. Anthropic says Claude asks before taking actions by default, while people who prefer a more continuous workflow can adjust how often it checks in. The idea is to provide more flexibility without removing the user’s ability to review, edit and approve the finished work.
Claude Projects are changing too
Anthropic followed the Docs and Slides announcement with another update on September 17, redesigning Projects in Claude Code. Although this update is aimed more directly at software development, it reflects the same broader idea: the user describes a goal while Claude organizes more of the work required to reach it.
The new Claude Code Projects experience can break a larger request into separate pieces, coordinate parallel work, review results and assemble them into a finished outcome. Anthropic says users can continue guiding the project while it progresses and can check in even when they are away from their main computer. The redesigned Projects experience is currently available in beta.
This is separate from Claude Docs and Slides, but the timing is notable. Across both announcements, Anthropic is moving from interfaces that primarily wait for individual instructions toward systems that can keep context, organize multiple steps and produce more complete results.
For everyday users, that shift is most visible in Docs and Slides. For developers, the redesigned Claude Code Projects experience applies a similar philosophy to larger software tasks.
Collaboration is becoming part of the AI workflow
AI-generated work often needs to be reviewed by other people, which makes collaboration an important part of any productivity platform. Anthropic is building that idea into the new Claude experience rather than assuming that the assistant works alone with a single user.
Claude Docs can support direct editing and comments, allowing people to make changes themselves or ask Claude for revisions around specific parts of the document. A colleague can provide feedback while Claude remains available to update the material, creating a workflow in which human collaboration and AI assistance happen around the same piece of work.
The same principle extends to presentations. Users can adjust slides directly instead of treating Claude’s first version as final, and the finished material can be shared or downloaded into familiar formats. That matters because real workplace documents rarely move from first draft to final version without revision.
Rather than replacing editing, the new tools appear designed to make editing more continuous. Claude can produce the first version, users can refine it, colleagues can add feedback and Claude can return to the same work with the surrounding context still available.
The rollout begins with paid Claude users
The unified Claude experience is not arriving for everyone at once. Anthropic says the merged Chat and Cowork experience is rolling out first to Pro and Max subscribers across the Claude apps on web, desktop and mobile, with additional plans expected to follow. Claude Docs, Claude Slides and Claude Design are being introduced in beta on paid plans, while Enterprise administrators will have control over when some of the new capabilities become available inside their organizations.
Anthropic says existing Cowork users will not need to rebuild their work. Their chats, projects, artifacts, connectors and skills remain available as the interface changes. People who primarily use Claude as a normal chatbot can also continue doing so without learning a completely different workflow.
That gradual rollout matters because the new features may not appear at exactly the same time for every account. A Pro or Max subscriber might receive the unified experience before someone on another plan, and organizational accounts can follow their own rollout schedule.
The mobile experience also plays a role in Anthropic’s broader vision. The company emphasizes that work created through Docs, Slides and Design can remain accessible across devices, making it possible to check progress or review something away from the desktop where the task originally began.
The real competition is moving beyond the chatbot
The most interesting part of this announcement may be what it says about the direction of the wider AI industry. The first major wave of generative AI products was built around chat windows. Users asked questions, received answers and copied useful parts somewhere else. The next stage is increasingly about what happens after the answer.
Reuters described Anthropic’s move as part of a wider competition to make AI assistants useful to enterprise and non-technical users by bringing more productivity functions directly into the assistant. Claude Docs and Slides make that shift especially visible because documents and presentations remain two of the most familiar forms of everyday professional work.
The question for users is gradually changing. Instead of asking only, “Can the AI write this?”, the more useful question may become, “Can the AI help me finish this?” Writing a paragraph is only one part of preparing a report. A finished project might require research, structure, editing, charts, presentation slides, review and sharing.
That is the space Anthropic is trying to occupy with the new Claude experience. The company wants the conversation to remain useful even as the work moves beyond conversation.
Documents and presentations make Claude feel more like a destination
There is a practical reason this approach could appeal to users: every switch between applications creates a little extra work. Copying text into a document is easy, but formatting it, preserving context, rebuilding charts and turning it into presentation slides all add additional steps. None of those steps is especially difficult by itself, yet together they can take a significant amount of time.
Claude Docs and Slides attempt to keep more of that process in one place. A document can develop alongside the research that created it. A presentation can draw from the document without requiring the user to manually rebuild the story. Design elements can be created from the same project context, and longer assignments can continue while the user focuses on something else.
Traditional productivity software is not suddenly unnecessary because Claude can create documents or presentations. Word processors, presentation tools and collaboration platforms have years of specialized features and established workflows behind them. What is changing is the role the AI assistant can play before the work reaches those applications.
Claude can increasingly become the place where a project is first understood, organized and developed, even when the final file eventually moves elsewhere.
Anthropic is reducing the gap between asking and creating
The broader theme connecting these updates is that Anthropic wants fewer steps between describing an idea and producing something usable.
A user can begin with a conversation about a topic, ask Claude to gather and organize relevant information, develop that material into a document and then prepare a presentation for a different audience. The same project can remain connected to previous conversations, available connectors and the context Claude has already gathered. If the work becomes larger, Claude can continue processing it without requiring the user to supervise every individual step.
For someone preparing a proposal, report, lesson, presentation or recurring update, that can make AI feel less like an additional tool alongside the normal workflow and more like part of the workflow itself.
There will still be reasons to move work into Microsoft Word, Google Docs or PowerPoint, and Anthropic supports those kinds of exports. The difference is that users may increasingly arrive in those applications with a much more complete first version rather than starting from a blank page.
That is why Claude Docs and Slides are more interesting than their names initially suggest. They are not simply two more buttons added to an AI assistant. They are part of Anthropic’s attempt to turn Claude into a place where work can move from an early idea to a shareable result without losing the conversation, information and decisions that shaped it along the way.
For Claude, the transition from chatbot to workspace is becoming much easier to see.
