Anthropic’s Claude Tag learns your company, one Slack message at a time

Anthropic introduces Claude Tag in the search preview, an “always-on” Claude who lives in Slack and works as an AI teammate. The new feature — which allows users to tag @Claude to provide insights in chats and assign tasks — will begin previewing search, and is available through Slack for Claude Enterprise and Claude Team customers.

Claude Tag is an evolution of many integrations that already exist. Users can already direct message @Claude within Slack or tag him in channels for on-demand help, and Claude Code in Slack routes coding tasks from channel mentions to full coding sessions on the web, posting updates back into the thread.

But Cloud Tag adds a layer of ongoing context and memory that is difficult to maintain with previous tools. “As Claude continues his channel, he learns more about the business,” a statement from Anthropic said. “Claude can also automatically collect facts from elsewhere in the organization, if he is given permission to read other channels.”

With a Claude Tag, everyone in a given Slack channel has access to a single Claude identity, meaning that “anyone can see what Claude has been working on, and can pick up the conversation where the last person left off.” System administrators will determine which tools, information, and channels Claude can access, and all of Claude’s identity will remain limited to the channels designated by administrators, so that a Claude set up for legal work cannot implant memories in the engineering channel, for example.

When assigned a specific task, Claude Tag will break the task down into phases and work through them using whatever tools he has access to, responding in a Slack thread with what he’s created. But Claude Tag also features an ambient mode that proactively jumps into the chat on its own to keep your team in the loop, tag things from across the organization, and catch up on forgotten topics or tasks.

Anthropic says this makes you feel like you’re “working with a real colleague – one who can produce work in front of an audience, with much more context and understanding than before.”

This context is an increasingly important part of enterprise deployments, and Anthropic isn’t the only company focusing on it. Microsoft also has Graph, which is expressed through Copilot and Work IQ. Snowflake and Databricks position their platforms as a back-end that contains tacit regulatory knowledge that agents can leverage. Glenn is also building an intelligence layer that understands the company context and sits between the model and enterprise data.