Anthropic launches Claude Tag, putting a shared AI teammate inside Slack

On June 23, Anthropic introduced Claude Tag, which it describes as "a new way for teams to work with Claude." Instead of each person chatting with their own assistant, Claude Tag drops a single shared Claude into a Slack channel that everyone can hand work to.

How it works

Anyone in the channel can mention @Claude and delegate a task in plain language. Claude breaks the request into stages, works through them with its connected tools, and replies in the thread with the result. The key difference from a typical chatbot is that it is built to be multiplayer: there is one Claude instance per channel, so the whole team can watch it work and pick up where a colleague left off.

"Claude Tag is built to be interactive and multiplayer. When Claude Tag works in a channel, everyone can see it, and everyone can jump in," Anthropic head of product Cat Wu told Fortune.

Claude Tag also builds context over time, remembering relevant information from the channels it sits in and the data sources it is connected to, so teams don't have to re-explain the same background on every request. It can plan tasks to run later and work asynchronously over hours or days rather than only responding in the moment.

If an organization turns on ambient behavior, Claude becomes proactive — flagging information it thinks you'll need from across its channels and tools, and following up on threads or tasks that have gone quiet without being resolved. Administrators can scope what Claude can see and do channel by channel, keeping sensitive data, tools, and memories tightly controlled.

Availability

Claude Tag is available today in beta for Claude Enterprise and Claude Team customers, running on the Opus 4.8 model, with launch credits going to eligible organizations. Anthropic says it plans to expand access over time.

Why it matters for builders

The headline figure for anyone learning to build with AI: Anthropic says that today "65% of our product team's code is created by our internal version of Claude Tag." That's a striking claim about how much day-to-day engineering an agent embedded in a team chat can absorb — and it lands as the industry shifts from single-user copilots toward agents that operate as shared, semi-autonomous teammates.

The launch also arrives as Anthropic's enterprise momentum builds. Fortune notes that, per Ramp's May AI Index, Anthropic edged past OpenAI in business adoption for the first time, with 34.4% of firms paying for its services against OpenAI's 32.3%.

For teams already routing standups, bug triage, and code questions through Slack, Claude Tag turns that channel into a place where work can actually get done rather than just discussed. The trade-off is the one every team now faces with capable agents: the more an AI teammate produces, the more disciplined your review and permission controls need to be. As Neowin reported, the beta leans on per-channel admin controls precisely so organizations can dial in how much autonomy that shared Claude gets.