Anthropic’s Claude Tag is learning your company, one Slack message at a time
TL;DR
Anthropic is bringing Claude Tag to Slack in beta for Claude Enterprise and Claude Team customers, letting employees mention @Claude in channels, assign work, and get threaded responses. The bigger shift is persistent channel memory: each Claude identity can keep context and work history so another teammate can continue from where the last person stopped. Admins define which channels, tools, and information each Claude instance can access, keeping, for example, legal context separate from engineering work.
Nauti's Take
The product story is framed as an AI teammate, but the real move is owning organizational context. That is the leverage point: models become more interchangeable when memory, permissions, and workflows live inside the system.
For teams, this can be useful if admin boundaries, transparency, and data hygiene are handled well. Without real governance, productivity quietly turns into a polite observer that knows too much.
Briefingshow
Claude Tag shows where enterprise AI is moving: away from a standalone chat box and toward a persistent context layer inside daily work. Whoever connects Slack history, tasks, tools, and internal knowledge fragments does not just answer questions; they starts shaping the company’s operational memory.
Sources