On June 23, 2026, Anthropic introduced Claude Tag, a new feature that embeds Claude as a team member inside Slack. Users can summon it with the @Claude tag to delegate tasks, and it is rolling out as a research preview (beta) for Claude Enterprise and Team plan customers.
June 23, 2026 · Anthropic
Claude Tag Lives Inside Your Slack — Watching, Learning, Acting
A new research preview puts a persistent Claude instance inside a channel, where it monitors activity and autonomously sends alerts, posts comments, and fixes code — learning a company's workflows one message at a time.
Shared
One Claude identity per channel — everyone keeps using it continuously
Ambient
Proactively joins talks, flags issues, follows up on forgotten tasks
Scoped
Memory is per-channel — a legal Claude can't touch engineering
The deepening Slack integration
Each stage adds more context, memory and autonomy.
Slack app
Existing · tag @Claude in DMs/threads
→
Claude Code
Dec 2025 · session from a bug thread
→
Claude Tag
Jun 2026 · channel-resident, proactive
How an ambient Claude works in a channel
MONITOR
Reads ongoing channel activity
→
LEARN
Builds memory of workflows
→
ACT
Alerts, comments, fixes code
Admins control which tools, information and channels Claude can access — part of a wider race against Microsoft Copilot, Glean and Snowflake/Databricks to feed in-house knowledge to AI.
Why teams like it
Less context switching and direct task delegation from conversations — automating everything from a bug report thread to PR creation.
The open concern
Privacy and security questions over how broadly Claude reads Slack data, and the risk of centrally managing corporate knowledge.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…