Anthropic launches reports to track companies’ AI spending — TechCrunch

Anthropic has introduced a new set of reporting tools for corporate clients using its artificial intelligence models. Managers will be able to see teams’ results over a selected period, token spending, work outcomes and areas where workflows failed, TechCrunch reports.

Assessing the cost of AI use

The reports will be generated based on internal analytics and employee surveys. They will also include recommendations for improving the results of using Anthropic’s tools. The company said it will not use this data to train models by default.

In addition, Anthropic is introducing shared group budgets so that teams can better control token spending. The new tools are intended to help clients assess the return on their AI spending and optimize it.

More current news is available on the UA.News Telegram channel Telegram.

Rising spending on coding tools

According to Jellyfish, which develops tools for programmers, Claude Code and other solutions based on Anthropic models remain the most widely used agentic tools for writing code among the company’s 276,000 clients whose data were included in the study. They accounted for nearly one-third of all proposed code changes submitted for review.

Jellyfish also recorded a tenfold increase in spending on such tools since the beginning of the year: their cost reached about 2% of total development spending. Jellyfish Chief Scientist Nicholas Arcolano noted that even with a relatively small share of spending, such growth was often not anticipated in chief financial officers’ plans.

According to data Anthropic shared with investors, the company recorded an operating profit in the second quarter. Net revenue retention was 900%: spending by the same clients in the third quarter of 2026 was nine times higher than in the third quarter of 2025.