Anthropic presents a research preview of Claude Tag, embedding Claude further into the enterprise Slack workflow. Unlike the previously called on-demand chat assistants, this function emphasizes continuous dialogue with the channel, building up team context within the mandate and working in a unified capacity.
Permanent Assistant for Slack
Claude Tag will first open the research preview to Slack users in Claude Enterprise and Claude Team. Users can get analysis, assign tasks directly from @Claude in the channel, or continue to work in the discussion thread.
Anthropic has previously provided various Slack integration methods, including private chats @Claude, ad hoc calls in channels, and Claude Code in Slack for encoded jobs to complete session on the webend. Claude Tag builds on these capabilities and adds ongoing context and shared memory.
Share the same identity in the channel
According to Anthropic, Claude Tag will keep up with the work being done by the team as the channel discusses. If the administrator grants authority, it can also extract relevant facts from other channels to supplement the information required for the current mission.
Within the same Slack channel, team members share the same Claude identity. This means that anyone can see what it has dealt with before and move on from where the previous dialogue stopped, without having to re-enter the background every time.
However, this memory capacity is not unlimited. System administrators can predetermine which tools, information and channels Claude can access. Each Claude identity will also be limited to a designated area, avoiding the mixing of context between different departments.
We'll dismantle the mission and follow it up.
When the user gives Claude Tag a clear task, it breaks the task down into multiple stages and calls for mandated tools to be completed gradually, and eventually returns the result in the Slack line.
In addition to reactive response, Claude Tag added a more proactive way of working. It can insert its own updates into the dialogue, alerting teams to cross-sectoral information or follow up on forgotten threads and tasks.
AI Competition to context
Such capabilities are becoming the focus of businesses when they deploy AI. Beyond the model itself, better access to intra-firm knowledge, competency systems and collaborative processes is affecting product competitiveness.
Currently, Microsoft is strengthening this direction through Copilot and Graph. Snowflake and Databricks are also emphasizing that their platform can serve as a base of business knowledge for AI proxy calls. Glenn is building the smart layer between the connection model and business data.