Anthropic’s new Claude Tag update lets its Slack agent read the full conversation — and jump in unprompted

Summary

Anthropic is betting that the biggest bottleneck in enterprise AI isn't model intelligence but the fact that most people still use AI alone. Its 'multiplayer AI' strategy aims to have agents operate across teams, read organizational context, and proactively insert themselves into work. The Claude Tag update now reads entire conversations, improving the accuracy of deciding when to jump in by about 30%.