Anthropic recently launched a new service called Claude Tag, positioned as an "always-on AI teammate" within Slack. Users can obtain insights or assign tasks simply by mentioning @Claude in chat; this feature is currently available in beta for Claude Enterprise and Claude Team customers. Claude Tag represents an upgrade built upon existing integrations rather than starting from scratch. Unlike previous one-to-one chats or on-demand mentions, it adds persistent context and memory capabilities—allowing Claude to continuously learn about team work through channel conversations, and, once authorized, automatically collect factual information from other channels within the enterprise. Each Slack channel shares a single Claude identity, enabling anyone to view Claude’s progress and take over where the previous user left off. Administrators configure which tools, data sources, and channels Claude can access—for instance, ensuring that a Claude configured for legal matters cannot inject memories into engineering channels, thereby maintaining clear permission boundaries. In Task Mode, Claude Tag automatically breaks down assignments into multiple stages, leverages existing tools to complete them step-by-step, and posts results back into Slack threads. It also features Environment Awareness mode, allowing it to proactively intervene in chats, synchronize team updates, flag important cross-organizational information, and even follow up on forgotten threads or tasks. According to Anthropic, this enables teams to experience “collaborating with real colleagues”—producing work in public settings while possessing far greater understanding power than before. Meanwhile, Microsoft’s Graph Copilot, Snowflake and Databricks’ enterprise knowledge platforms, and Glean’s intelligence layer are all pursuing similar directions, making competition in enterprise-grade AI context increasingly fierce.