Anthropic has launched Claude Tag, a new product that replaces its Slack app with a persistent AI teammate designed to learn a company's context, monitor its channels, and work autonomously, according to VentureBeat and TechCrunch. Where the old integration responded when called, Claude Tag is meant to be always present -- accumulating knowledge about a team's projects, people and norms over time.
The shift is conceptual as much as technical. By learning a company 'one Slack message at a time,' Claude Tag aims to become more useful the longer it operates, building up institutional memory that a stateless chatbot can't match. That persistence is also a competitive moat: an agent that deeply understands a specific organization is harder to rip out and replace.
“That persistence is also a competitive moat: an agent that deeply understands a specific organization is harder to rip out and replace.”
The autonomy cuts both ways. An AI teammate that monitors channels and takes actions on its own promises real productivity gains but raises governance questions about oversight, permissions and trust -- especially in sensitive workplace communications. The launch sharpens Anthropic's enterprise positioning against OpenAI and Microsoft, all racing to define what an embedded AI coworker actually looks like.