News in Short

• Anthropic has launched Claude Tag in beta for Claude Enterprise and Team customers.
• Claude Tag allows teams to tag @Claude directly in Slack channels and assign tasks.
• The AI can remember channel context, use connected tools, and work asynchronously.
• Anthropic says 65% of its product team’s code is now created by its internal version of Claude Tag.

Anthropic has introduced Claude Tag, a new AI collaboration feature that turns Claude into a shared team member inside Slack. The tool allows employees to tag @Claude in conversations, assign tasks, and let the AI work independently while building context from approved channels, tools, data sources, and codebases.

The launch marks a significant step in Anthropic’s effort to move AI beyond one-on-one chat interactions. Instead of individual conversations, Claude Tag is designed to function as a collaborative teammate that multiple people can work with inside a shared workspace.

The feature is available in beta starting today for Claude Enterprise and Team customers.

What Is Claude Tag and How Does It Work?

Claude Tag is Anthropic’s latest workplace AI tool built around team collaboration. It integrates directly with Slack and allows organizations to add Claude as a member of selected channels.

Once access permissions are configured, employees can simply mention @Claude in a channel and assign work using natural language. Claude then breaks the task into smaller steps, accesses approved tools and information, and completes the assignment before responding in a Slack thread.

Unlike traditional chatbots that depend on continuous prompting, Claude Tag is designed to work independently after receiving instructions. Employees can move on to other priorities while the AI completes the task.

Anthropic says this approach mirrors how teams delegate work to colleagues rather than constantly managing a conversation.

How Is It Different From Traditional AI Assistants?

The biggest difference is that Claude Tag operates as a shared AI identity rather than a personal assistant.

According to Anthropic, each Slack channel has one Claude instance that everyone can interact with. This means conversations, context, and ongoing work remain visible to the team.

As a result, users can join discussions already in progress and continue work without restarting the interaction. The company believes this creates a more collaborative experience than traditional AI chat windows.

Claude Tag also remembers information from the channels where it has permission to operate. Over time, it develops a deeper understanding of ongoing projects, team discussions, and workflows.

This persistent context reduces the need for repeated explanations and allows the model to provide more relevant responses.

Can Claude Tag Learn and Take Initiative?

Anthropic says Claude Tag can gradually build knowledge from approved channels and connected data sources.

The company describes this as acquiring “tacit knowledge” about a team’s work. That context helps the AI provide more informed answers and execute tasks with greater accuracy.

Another notable capability is proactive behavior.

When the optional ambient mode is enabled, Claude can surface updates it believes are relevant. It can monitor connected information sources, identify developments that may require attention, and follow up on unresolved discussions.

The AI can also notify teams when conversations appear stalled or when tasks remain incomplete.

Anthropic positions this feature as a way to reduce missed information and improve visibility across projects.

How Does Claude Tag Support Asynchronous Work?

One of the core ideas behind Claude Tag is asynchronous collaboration.

Instead of requiring employees to remain actively involved, Claude can continue working after receiving a request. It can even schedule future actions and revisit tasks over hours or days.

Anthropic says this has become a common workflow internally. Teams often delegate multiple assignments to different Claude instances simultaneously while focusing on higher-priority work.

The company claims the approach has already become a significant part of its own operations.

According to Anthropic, around 65% of its product team’s code is now generated using its internal version of Claude Tag.

Beyond software development, employees reportedly use it for support investigations, product analytics, debugging, and operational tasks.

What Controls Are Available for Enterprises?

Because Claude Tag can access business data and internal tools, Anthropic has built the product around administrative controls.

System administrators determine which channels Claude can access and which tools, databases, or code repositories it can use. Separate Claude environments can be created for different departments.

For example, a sales-focused Claude can operate with sales-related tools and information while remaining isolated from engineering systems. Anthropic says memories and learned context remain confined to the channels assigned to each configuration.

Administrators can also set spending limits, review activity logs, and track who requested specific tasks. These controls are intended to help organizations maintain oversight while expanding AI usage across teams.

What Does Claude Tag Mean for Workplace AI?

The launch reflects a broader shift in how AI is being deployed inside organizations.

Rather than acting solely as a personal assistant, AI is increasingly becoming a shared resource that teams can collaborate with collectively. Claude Tag pushes that concept further by combining memory, tool usage, proactive updates, and asynchronous execution within existing workplace software.

For Anthropic, the goal appears to be making AI a persistent part of everyday team workflows instead of a separate application employees open when needed.

As enterprise adoption of AI accelerates, Claude Tag could become an early example of how organizations integrate AI agents directly into collaborative work environments. With Slack serving as its first platform, Anthropic’s new Claude Tag signals a growing move toward AI teammates that work alongside entire teams rather than individual users.