Anthropic says Claude now handles 95% of its internal analytics queries, according to InfoQ. The figure points to a mature internal deployment where employees can ask business-data questions without relying on a traditional analyst workflow for every request.
The claim matters because analytics is one of the clearest enterprise use cases for language models: users want answers from data, but most organizations still struggle with fragmented tools, permissions, and semantic definitions.
If the system is reliable, it shows how AI assistants can become a front end for operational knowledge. The hard parts remain governance, metric consistency, and knowing when a human analyst needs to review the result.