Anthropic's Claude Now a Persistent Slack Teammate

Anthropic has introduced Claude Tag, embedding its most advanced AI model into Slack as a persistent, shared teammate. This product, available for Enterprise and Team customers, represents a significant move to integrate AI directly into the enterprise collaboration layer.

Claude Tag functions as a standing member of a team, capable of building memory, taking initiative, and working asynchronously. It operates within Slack channels, responding to user requests by executing tasks using integrated tools and providing results in a thread. The system runs on the Claude Opus 4.8 model.

Key differentiating features include its multiplayer nature, where one Claude instance interacts with everyone in a channel, and its ability to learn over time by accumulating context. Claude Tag can also take initiative by proactively surfacing relevant information and following up on unresolved tasks.

Enterprise security and governance are central, with administrators able to define separate Claude identities for specific channels, tools, and data access. This isolation ensures that memories and data remain within defined boundaries. Administrators can also set token-spend limits and review logs of all Claude's actions.

The launch places Anthropic in direct competition with other AI players integrating agents into Slack, such as Salesforce, OpenAI, and Perplexity. The company highlights the significant growth of Claude Code, with enterprise use comprising over half of its revenue, underscoring the strategic importance of channel-level AI presence.

Risks for enterprise buyers include vendor dependency due to deep integration, governance challenges with ambient monitoring, unclear pricing for continuous usage, and potential reliability concerns given Anthropic's recent infrastructure strain. Claude Tag signals a future where AI acts less like a tool and more like a new category of worker.