Anthropic Launches Claude Tag for Slack Teams

Anthropic on Tuesday launched Claude Tag, a Slack-native AI agent that replaces the company’s existing Claude in Slack app and functions as a persistent, shared team member. Available immediately in beta for Claude Enterprise and Team customers, the product lets any Slack user delegate tasks by typing @Claude in any channel.

What Claude Tag Does in Slack

Claude Tag embeds Anthropic’s most advanced model directly into Slack as a standing team member — not a single-user chatbot — that builds shared memory, takes initiative across channels, and works asynchronously without requiring a dedicated prompt session. Any team member can @mention it in any channel to assign tasks, request analysis, or generate drafts. According to Anthropic’s announcement, Claude Tag can monitor channels, surface relevant information proactively, and interact with multiple people simultaneously rather than serving one user at a time.

The distinction from a conventional AI assistant is structural. Claude Tag is designed to accumulate institutional knowledge over time — tracking decisions, assignments, and context that build up inside Slack channels — and act on that knowledge without waiting to be explicitly prompted.

Anthropic’s Internal Usage as Proof of Concept

Anthropically’s most concrete claim for Claude Tag is its own internal adoption: the company says 65% of its product team’s code is now generated by an internal version of Claude Tag, according to VentureBeat’s coverage of the launch. Anthropic also reported running internal support channels and data insight workflows through the same system.

The figures are self-reported and unaudited, but the assertion is specific: Anthropic is stating that the majority of its own product engineering output already flows through the tool it just released to customers. That framing — dogfooding at scale — is central to how the company is positioning Claude Tag’s enterprise credibility.

Availability and Pricing Tier

Claude Tag launched in beta on Tuesday and is available exclusively to Claude Enterprise and Claude Team subscribers. It is not available on individual or free tiers. The product directly replaces Anthropic’s prior Claude in Slack integration, meaning existing Enterprise and Team customers in Slack will see the transition to Claude Tag rather than running both products in parallel.

Anthropics has not disclosed specific pricing changes tied to the Claude Tag launch beyond the existing Enterprise and Team plan structures.

How Claude Tag Differs from Prior AI Slack Integrations

Most AI tools embedded in Slack — including Anthropic’s previous integration — function as single-user assistants: one person opens a DM or mentions the bot, asks a question, and gets a reply. Claude Tag is architected differently. According to Anthropic’s product description, it operates at the channel level, meaning it can observe ongoing conversations, build context from multiple participants, and respond to or initiate interactions with any member of that channel.

This positions Claude Tag closer to an autonomous agent than a retrieval tool. It is designed to take initiative — flagging relevant context, suggesting next steps, or completing delegated tasks — without waiting for an explicit prompt. That behavioral model raises distinct questions around oversight, since the agent can act on behalf of a team rather than a single accountable user.

Enterprise Governance Implications

For IT and security leaders evaluating Claude Tag, the shared-memory and autonomous-action model introduces governance considerations that single-user AI tools do not. Because Claude Tag builds persistent context across an entire Slack channel — potentially including sensitive business decisions, personnel discussions, or unreleased product details — data residency, retention policies, and access controls become more complex.

Anthropics has not published detailed documentation on how Claude Tag’s channel-level memory is stored, who can audit it, or how it interacts with existing Slack Enterprise Grid data governance features. Enterprise buyers evaluating the tool will need to press those questions before broad deployment, particularly in regulated industries.

What This Means

Claude Tag is Anthropic’s clearest attempt yet to own a structural layer of enterprise work rather than sit at its edges. By embedding an AI agent at the channel level — where decisions are made and institutional knowledge accumulates in real time — Anthropic is competing not just with other AI assistants but with the collaboration layer itself.

The 65% internal code-generation figure, if accurate, signals genuine internal conviction. But the product’s autonomous, channel-wide scope also means enterprises will need clearer answers on memory governance, auditability, and role-based access before treating Claude Tag as a default team member rather than an opt-in utility. The beta label gives Anthropic room to refine those controls; enterprise buyers should treat it as such.

FAQ

What is Claude Tag?

Claude Tag is a Slack-native AI agent from Anthropic that functions as a shared team member across channels, not a single-user chatbot. It builds persistent memory, works asynchronously, and can be addressed by any team member using @Claude.

Who can access Claude Tag?

Claude Tag is available in beta to Claude Enterprise and Claude Team subscribers. It replaces Anthropic’s previous Claude in Slack app and is not available on free or individual plans.

How does Claude Tag differ from a standard AI chatbot in Slack?

Unlike single-user AI integrations, Claude Tag operates at the channel level — observing conversations across multiple participants, accumulating shared context over time, and taking initiative without requiring an explicit prompt from a single user.