Wednesday Jun 24

Claude Becomes A Slack Teammate

23JUN
@CLAUDETEAM CHATCLAUDE

Anthropic put Claude inside Slack as a shared coworker. Type @Claude in any channel and it runs multi-step tasks, then hands them off mid-thread. One shared AI identity for the whole team.

This reframes the assistant. Instead of a private helper, you get a coworker the group manages together.

It plans a job, executes the steps, and reports back. Colleagues resume each other's unfinished work. Admins lock down which systems and memories it may reach.

Rolling out to paid business plans first. The bet is software you delegate to, not just chat with. Open question: does pooled memory aid the group or spill secrets between squads.

full brief & sources

Why this matters

  • First Anthropic product that frames Claude as a shared coworker, not a personal assistant.
  • Slack is where work coordination already happens. Claude now sits inside it.
  • Shared identity plus handoff changes how teams split work with AI.

🔍 What happened

  • Announced June 23, 2026. Live for Claude Enterprise and Team customers.
  • Summon it with @Claude in any channel.
  • Breaks tasks into stages, runs them independently, then delivers in Slack.
  • An 'ambient' mode lets it follow up on threads and tasks on its own.
  • Admins tightly scope tools, data, and memory per channel.

💬 Smart takes

  • Bloomberg: Anthropic wants Claude to be your new Slack coworker.
  • Engadget: 'Sorry, Slackbot. Claude is taking your job.'
  • Skeptic: a shared AI identity with broad channel access is a data-governance headache waiting to happen.

🧭 Where this goes

  1. LikelyOpenAI and Google ship rival in-Slack or in-Teams agents within 90 days.
  2. Likelyenterprises pilot Claude Tag in one team before any company-wide rollout.
  3. Possible'shared AI identity' becomes a standard enterprise pattern, not a feature.
  4. Wild CardSlack restricts third-party agents to push its own Agentforce-style layer.

🥄 The Spoon Take

The assistant era is ending. The teammate era is starting. Anthropic is not selling a smarter bot. It is selling a seat on the team that never logs off. The fight now is over where the AI sits, not how smart it is.

🤔 Pushback

A shared identity everyone can summon and that follows threads on its own is one bad scope away from leaking the wrong data to the wrong channel.