June 24, 2026. On June 23, Anthropic launched Claude Tag, a way for teams to delegate work to its Claude AI by tagging it in their group chat. It starts on Slack. An administrator connects Claude to chosen channels, tools, data, and even codebases, and from then on anyone in a channel can tag @Claude with a request and hand off the task while they move on to other work. Anthropic frames it as an evolution of its Claude Code product, and says 65 percent of its own product team's code is now created by an internal version of the tool, with the same pattern spreading to data pulls, support tickets, and debugging.
What is new
- It is multiplayer. There is one Claude per channel that everyone can see and pick up from, more like a teammate than a private chatbot.
- It learns over time. It builds context from the channels it is in, and does not read private channels, so the team stops re-explaining the basics.
- It takes initiative. With an ambient mode enabled, it flags information it thinks you need and follows up on threads that have gone quiet.
- It works asynchronously. You can set it a task or let it schedule its own work, pursuing a project over hours or days.
The guardrails
Access is scoped per channel, so a Claude set up for sales does not share memory or tools with one set up for engineering. Administrators set token spend limits for the organization and for individual channels, and can view a full audit log of every task and who requested it. Claude Tag is in beta for Claude Enterprise and Team customers, runs on Opus 4.8, and replaces the older Claude in Slack app, with a 30 day window to migrate.
What it means for operators
The brand matters less than the pattern. AI agents are moving into the chat tools teams already live in, which is exactly where repeatable work, research, data lookups, first pass support replies, and status chasing, tends to pile up. The same shift showed up days earlier when OpenAI pushed its Workspace Agents into business chat and Drive, which we covered in our June 22 brief. For an SMB or agency, the opportunity is to delegate that repeatable work where the work already happens, but the discipline is the same as any agent rollout: scope access tightly, keep an audit trail, and keep a human on anything irreversible, the governance point from our June 19 analysis. We build exactly these scoped, audited agent workflows through our AI automation and AI automation agency services, and you can hire an AI engineer to wire them into the tools your team already uses.
Frequently Asked Questions
It is a way to delegate work to Anthropic's Claude by tagging @Claude in your team's chat. It launched on June 23, 2026 starting with Slack, where an admin connects Claude to specific channels, tools, and data, and anyone in the channel can hand it a task.
It is built to act like a teammate. It is multiplayer (one shared Claude per channel), it learns context over time, it can take initiative with an ambient mode, and it works asynchronously, including scheduling its own tasks over hours or days.
It is in beta for Claude Enterprise and Team customers, works inside Slack, and runs on Claude Opus 4.8. It replaces the older Claude in Slack app, and administrators have a 30 day window to migrate.
The pattern is copyable with lighter tools: put an AI agent where your team already works and let it handle repeatable tasks. Apply the same guardrails Anthropic builds in, scope its access per use, set spend limits, keep an audit log, and keep a human on irreversible actions.