Anthropic launches Claude Tag: AI agent becomes a full-fledged team member in Slack
Anthropic has introduced a new tool, Claude Tag, which integrates a neural network into Slack work channels as an active project participant. This is not just a chatbot — Claude Tag can perform complex multi-step tasks using connected databases, program code, and third-party applications. To assign a task, simply mention the model via @Claude, and the AI independently breaks the assignment into stages, working in the background.
Key features of Claude Tag
- Multi-user mode: all channel participants can see Claude's actions and continue a conversation started by a colleague.
- Context accumulation: the neural network remembers discussion details and does not require repeated explanations.
- Proactivity: when background mode is enabled, the AI independently reports important updates or reminds about unresolved tasks.
- Autonomy: the model can plan and execute long-term projects over several days.
Within Anthropic itself, Claude Tag already creates 65% of program code. The tool is also actively used for analyzing metrics and processing support tickets. Administrators can restrict the neural network's access to confidential data and set limits on token usage.
Currently, the service is available in beta testing mode for users of the Claude Enterprise and Team plans. Claude Tag operates on the Opus 4.8 model and replaces the standard Claude application for Slack.
Expert comment: Integrating AI agents into communication platforms is a logical step in the evolution of work processes. Claude Tag demonstrates how neural networks can not only respond to queries but also take over project management and routine tasks. However, it is worth noting that the autonomy of such agents requires careful control by administrators, especially in the context of data security and compliance with corporate policies. In the coming years, we will likely see widespread adoption of such solutions, which will fundamentally change the approach to teamwork.