On June 23, 2026, Anthropic announced Claude Tag, a new configuration of its Claude assistant designed to operate as a shared team member inside Slack. The feature is available in beta today for Claude Enterprise and Team customers.
What Claude Tag does
Claude Tag allows organizations to add Claude to selected Slack channels and grant it access to chosen tools, data sources, and even codebases. Anyone in a channel can tag @Claude to delegate tasks while other team members focus on different work. The assistant builds context by remembering relevant information from the channels it’s part of, breaks tasks into stages, and can schedule and execute multi-step work over hours or days.
Anthropic positions Claude Tag as an evolution of Claude Code: it makes the model more proactive and better suited for full-team workflows. Anthropic reports that, internally, 65% of its product team’s code is produced by its internal version of Claude Tag.
Key differences from single-user chatbots
-
Multiplayer interaction: within a given Slack channel there is a single Claude that interacts with everyone. This visibility lets any team member see what Claude is working on and pick up conversations where others left off, making the experience more like collaborating with a teammate than a one-off chat.
-
Learning over time: as Claude follows a channel, it accumulates contextual knowledge about the work. With permission, Claude can also learn automatically from other Slack channels and data sources (it does not report from private channels), which helps it provide higher-quality outputs.
-
Initiative: when “ambient” behavior is enabled, Claude will proactively surface updates, flag relevant information across its connected channels and tools, and follow up on quiet or unresolved threads and tasks.
-
Asynchronous work: teams can assign tasks to Claude and continue other priorities while Claude proceeds autonomously. Claude can also be messaged directly in private DMs, where it will use the personal tools and connectors configured for the user.
Administration, permissions and cost controls
Claude Tag is built for teams and organizations: administrators can tightly control which data and task-specific tools the model can access and in which channels. Administrators configure access so that separate Claude identities are effectively scoped to channel sets; memories and permissions remain restricted to those scopes. For example, a model configured for sales will not pass memories to one configured for engineering, nor expose sales tools to engineers.
Administrators can set token-spend limits at the organization and channel level, and they can view logs of everything @Claude has done along with the identity of the requester for each task.
Launch details and migration
Claude Tag’s beta is available starting June 23, 2026, to Claude Enterprise and Team customers. Setup includes pairing Claude Tag with a Slack workspace, granting access to tools, setting a monthly spend limit, and testing the assistant in a private channel.
Claude Tag replaces the existing Claude in Slack app; administrators can opt into migration within a 30-day window. Anthropic is issuing an introductory launch credit to eligible Enterprise and Team organizations so companies can try the new service.
Technically, Claude Tag runs with Opus 4.8. Further product documentation and details are available from Anthropic.
Why this matters
Claude Tag shifts the interaction model from single-user AI chats to a shared, team-aware assistant that can remember context across channels, act proactively, and be governed centrally with permissions, logging, and cost controls. Anthropic’s internal usage suggests the approach speeds up not only engineering work but also product metrics tracking, support ticket handling, and root-cause analysis of complex bugs, illustrating a broader move to embed AI as a collaborative team member in daily workflows.
