Anthropic introduced reporting features that combine enterprise AI usage records with employee surveys. Managers can examine work performed over a period, its outcomes and costs, where workflows stalled, and suggested improvements. Pooled team budgets add a way to manage token spending around work rather than only individual usage. The company says it does not use this data for model training by default. Reporting alone does not prove productivity gains or return on investment, so output quality and the organization’s evaluation criteria remain important.