anthropic.com via Hacker News June 23rd 2026

Anthropic Launches Claude Tag for Team Slack Collaboration

7 sources tracking this story

anthropic agents enterprise ai ai-product enterprise-ai agents

TL;DR

Claude Tag gives each Slack channel a single shared Claude identity that any team member can address, redirect, or hand off without re-briefing the agent.
The August 3 retirement of the legacy Slack app forces enterprise IT teams into an active migration decision with a hard deadline, not a passive evaluation.
TechCrunch identifies Microsoft Graph, Snowflake, Databricks, and Glean as the direct rivals Anthropic is targeting for enterprise institutional-knowledge capture.

Editor's note

Claude Tag is Anthropic's first product designed to accumulate organizational context across a team rather than serve a single user, embedding a shared agent identity inside Slack channels that any team member can address, redirect, or hand off to. The competitive landscape TechCrunch maps out (Microsoft Graph, Snowflake, Databricks, and Glean as direct rivals) signals this is a fight for the enterprise intelligence layer, not a Slack productivity add-on. Ramp's May 2026 AI Index shows Anthropic already leads OpenAI in enterprise adoption for the first time (34.4% vs. 32.3%), with Claude Code as the primary driver, and Claude Tag extends that lead into team workflows where institutional context compounds across conversations. The August 3 forced migration deadline for the legacy Slack app converts the launch from a beta announcement into an active IT procurement decision, and Anthropic's own 65% internal code-generation rate through Claude Tag functions as the product's strongest proof point.

Claude Tag, Anthropic's new Slack integration, is built around a different model than the individual chatbot session most practitioners are used to. Teams grant Claude access to selected channels, data sources, tools, and codebases, then @-mention it to delegate tasks. Claude builds context by following channel discussions over time and can work asynchronously, handling assignments that stretch across multiple days without requiring the original requester to stay engaged.

The multiplayer design is the notable architectural choice. One Claude instance serves the entire channel, meaning teammates can see what Claude is doing, continue its work, and build on prior context together. When ambient behavior is enabled, Claude also proactively flags relevant information and follows up on unresolved threads without being explicitly asked.

Anthropics points to internal adoption as validation: the company says 65% of their product team's code is generated by their internal version of this technology. Described use cases extend beyond engineering to metrics analysis, support ticket management, and debugging assistance.

The beta is available to Claude Enterprise and Team customers on Slack, runs on Claude Opus 4.8, and includes administrative controls covering data access, token spending limits, and activity logging. Eligible organizations receive launch credits for trial usage.

What the announcement does not address is how retained channel context is managed over time: how long Claude holds that conversation history, under what conditions it is cleared, and what that means for organizations with sensitive communications running through Slack. For any team seriously evaluating ambient mode, those data retention questions are the ones to resolve before enabling it.

What others are reporting

Coverage cluster as of 24h after publish

Bloomberg Read →

Tier-1 business press validation of the launch; Bloomberg's framing positions Claude Tag as a workplace identity play rather than a feature release.
TechCrunch Read →

Names Microsoft Graph, Snowflake, Databricks, and Glean as rivals; frames Claude Tag as a contest for enterprise institutional knowledge, not task automation alone.

As Claude follows along with its channel, it learns ever more about the work. Claude can also automatically gather facts from elsewhere in the organization.
Fortune Read →

Adds enterprise adoption data from Ramp's May AI Index (Anthropic 34.4%, OpenAI 32.3%) and frames the launch within Anthropic's pre-IPO enterprise revenue push.

Claude Tag is built to be interactive and multiplayer. When Claude Tag works in a channel, everyone can see it.
VentureBeat Read →

Headline foregrounds the replacement framing, centering the August 3 migration deadline and the product shift from on-demand integration to permanent team agent.
Artificial Intelligence News Read →

Balances the launch against governance risks: misconfigured access boundaries, autonomous execution hazards, and compliance overhead for regulated industries.

Instead of a private back-and-forth, Claude Tag shows up in the open. (Rob Seaman, Slack General Manager)
NewsBytes Read →

Details the security architecture: isolated Claude identity per channel, scoped admin controls, private channels explicitly off-limits by design.

65% of its product team's code is now generated by this internal version of Claude Tag.

Shared on Bluesky by 4 AI experts

Originally reported by anthropic.com

Read the original article →

Original headline: Anthropic Launches Claude Tag: Agentic AI Coworker for Slack That Builds Context, Delegates Multi-Step Tasks, and Works Asynchronously — Beta for Team and Enterprise