Anthropic has introduced new reporting tools intended to help businesses understand how employees use its artificial intelligence products and what those services cost. TechCrunch reported on October 28 that the tools use internal analytics and employee surveys to give managers a view of team activity over a selected date range. The reports cover work produced, associated spending and points where workflows break down.
Anthropic also said the reports would recommend ways to improve results and that the underlying data would not be used for model training by default. The company is introducing pooled group budgets as another way for teams to manage their spending on tokens, the units used to process text and other information through AI models.
The announcement addresses a practical problem for companies adopting AI: usage can increase quickly, but managers need evidence that the expense is worthwhile. TechCrunch cited a report from software-development tools company Jellyfish covering 276,000 customers. That report found that coding tools using Anthropic models accounted for nearly a third of proposed code changes submitted for review. Spending on these tools had grown tenfold since the beginning of the year and represented about 2% of total engineering costs, according to TechCrunch.
For business teams, the new reporting approach could make it easier to compare AI expenditure with work delivered and identify inefficient processes. However, usage and cost reports do not by themselves prove that a tool improves productivity. The announcement reflects a broader enterprise need to measure results as AI systems become part of regular workplace operations.