Anthropic’s Claude Tag Turns Slack Into An Always-On AI Teammate

Anthropic is bringing Claude deeper into workplace collaboration with Claude Tag, a new beta service that lets teams bring @Claude into Slack channels, assign tasks, and let the AI (artificial intelligence) work with shared organizational context.
 

TL;DR

• Claude Tag is now available in beta for Claude Enterprise and Team customers.
• Teams can tag @Claude in Slack channels, direct messages, or the AI assistant panel.
• Claude can remember channel context, use approved tools, and work asynchronously.
• Admins can control access, spending, logs, and channel-specific Claude identities.

Anthropic Says Claude Tag Makes Slack A Home For AI Teamwork

Anthropic has introduced Claude Tag, a new service designed to make Claude feel less like a standalone chatbot and more like an AI teammate inside Slack.

The feature allows users to tag @Claude in a Slack conversation, hand over a task, and follow its work in the thread. Claude can break requests into stages, use the tools it has permission to access, and return completed work directly where the conversation is happening.

The beta is available for Claude Enterprise and Claude Team customers, with Anthropic starting in Slack because it sees the workplace messaging platform as a natural space for team collaboration with AI.
 

How Claude Tag Uses Shared Slack Context To Work Like A Teammate

Unlike one-off AI chats, Claude Tag is built around a shared Claude identity inside a Slack channel. This means everyone in that channel can see what Claude is doing, continue the conversation, or steer the task from where another teammate left off.

Claude can also build context over time by remembering relevant information from the channels it has access to. That reduces the need for employees to repeatedly explain background details, while giving Claude more of the tacit knowledge needed to support ongoing work.

Anthropic says Claude can be connected to selected tools, data, and even codebases. In its internal use, the company says 65% of its product team’s code is created by its internal version of Claude Tag.
 

Claude Tag Can Take Initiative, Follow Up, And Work Asynchronously

Claude Tag is not limited to responding when prompted. If ambient behavior is enabled, Claude can proactively flag relevant information, post updates, and follow up on threads or tasks that have gone quiet without being resolved.

It can also work asynchronously, allowing employees to assign a task and return to other priorities while Claude continues working. Anthropic says Claude can even schedule tasks for itself and pursue projects over hours or days.

Beyond channel tagging, users can also message Claude directly or open the Claude panel inside Slack. In private surfaces, Claude uses the personal tools and connectors enabled in the user’s own Claude account, while channel tagging uses the tools and access configured by the organization.
 

Anthropic Gives Admins Tight Controls Over Claude Tag Access

For businesses, the biggest question may be control, and Anthropic is positioning Claude Tag around administrator-led governance.

System administrators can decide which tools, information, and channels Claude can access. Each Claude identity remains scoped to the channels defined by admins, meaning a Claude configured for sales work will not carry memories into an engineering setup or expose sales tools to unrelated teams.

Admins can also set token spend limits for the full organization and individual channels. They can view logs showing what @Claude has done and who requested each task.

Claude Tag replaces the existing Claude in Slack app. Administrators can opt in within 30 days, and eligible Enterprise and Team organizations will receive introductory launch credits. Anthropic says Claude Tag works with Opus 4.8.

For enterprises trying to move AI from side-window assistant to everyday workflow participant, Claude Tag is a clear signal of where Anthropic wants Claude to sit next: inside the team conversation, not outside it.