AI is moving from sidekick to teammate
The current way we use AI is largely solitary—a private chat where you ask a question and get an answer. But for a professional team, that model is too disconnected from how work actually happens.
Anthropic is trying to change that with Claude Tag. By moving Claude into Slack channels as a "multiplayer" participant, they are attempting to turn the AI from a private assistant into a visible teammate.
The idea is to tag @Claude in a channel to delegate a task, then walk away. It can work asynchronously, planning and executing projects over hours or even days using Opus 4.8. Because it lives in the channel, it builds context from the conversation, meaning you don't have to re-explain the same background info every time you ping it.
This isn't just a theory for Anthropic; they claim 65% of their product team’s code is already being written by their internal version of this tool.
To keep it from becoming a security mess, the system lets admins scope Claude’s "memories" and tool access to specific channels. This ensures an engineering bot stays in its lane and doesn't accidentally ingest sensitive data from a sales channel.
AI is shifting from a tool you use privately to a visible, asynchronous collaborator that operates directly within your existing team workflows.