Anthropic Claude Tag: Slack always-on AI teammate preview

Anthropic is introducing Claude Tag in research preview, an always-on Claude that lives in Slack and acts as an AI teammate. The feature will begin in research preview and is available through Slack for Claude Enterprise and Claude Team customers, the company announced on November 4.

How does Claude Tag work?

Claude Tag is an always-on Claude that follows a Slack channel, maintains persistent context, and responds in threads or proactively when appropriate. Anthropic says “As Claude follows along with its channel, it learns ever more about the work,” and that Claude can automatically gather facts from elsewhere in the organization if it is granted permission to read other channels.

When assigned a task, Claude Tag breaks the work into stages and uses whichever tools it has access to, responding in a Slack thread with what it has created. Administrators specify which tools, information, and channels each Claude identity can access, and each Claude identity is scoped to the channels the admins define so memories remain channel-specific.

How does Claude Tag differ from existing Slack integrations?

Claude Tag builds on earlier Slack integrations but adds persistent context and an ambient presence that previous tools did not maintain continuously. Users could already DM @Claude or tag it for on-demand help, and Claude Code in Slack routes coding tasks from channel mentions to full coding sessions on the web, posting updates back into the thread. Claude Tag differs by providing a single Claude identity per channel that anyone in that channel can access and continue a conversation from the last person who left off.

Claude Tag also features an ambient mode that can proactively jump into chats to keep teams updated, flag items from across the organization, and follow up on threads or forgotten tasks. Administrators control scope so, for example, a Claude set up for legal work cannot seed memories into an engineering channel.

Why it matters

Persistent, scoped context is increasingly central to enterprise AI deployments because it lets an assistant accumulate situational awareness across conversations instead of treating each mention as stateless. Anthropic positions Claude Tag as a way to make the assistant operate more like a team member, with shared visibility and continuity: Anthropic says it can feel like “working with a real colleague — one that can produce work in public view, with far greater context and understanding than before.”

Other companies are pursuing adjacent paths to the same problem: Microsoft expresses organizational context through Graph, Copilot, and Work IQ; Snowflake and Databricks pitch their platforms as back-end stores of tacit company knowledge for agents to tap; Glean is building an intelligence layer to sit between models and enterprise data. Claude Tag joins this set of efforts by focusing on an always-on, channel-scoped identity inside Slack.

What to watch

Watch which customers enter the research preview and how administrators configure channel and tool access; those settings will determine how broadly Claude Tag can surface organizational facts. Also watch whether ambient interventions and cross-channel fact gathering trigger new governance or privacy controls inside organizations.