Friday, Oct 30 | --:--
Back to home

Anthropic Shows Claude Enterprise Admins What Each Team's AI Spend Produced, and Lets Them Pool Budgets

Anthropic made smart reports, in-app surveys and pooled group budgets generally available for Claude Enterprise on October 28. The reports summarise a sample of transcripts to show workstreams, spend and friction; Anthropic says they are not for individual performance reviews. TechCrunch cited Jellyfish data showing spending on AI coding tools up tenfold this year.

Times of AI Desk 3 min read San Francisco View as Markdown
Four small glass jars pour glowing liquid into one large shared jar on a table, a blank card lying beside it.

Anthropic's newest enterprise features answer the question finance teams are now asking about AI tools: what did the spend buy? On October 28 the company made generally available three Claude Enterprise admin tools: smart reports, in-app surveys and pooled group budgets.

What admins get#

Smart reports analyse a sample of an organisation's Claude transcripts to summarise which workstreams teams use Claude for, what that costs and where users hit friction, according to Anthropic. In-app surveys let admins ask users questions inside Claude. Pooled budgets let a group share one spending limit, with alerts at 50%, 75%, 95% and 100% of the budget.

Anthropic says the reports are not meant for judging individual performance or making employment decisions. TechCrunch reported that Anthropic will not use the data to train its models by default. Availability varies for customers on HIPAA or customer-managed encryption key setups, Anthropic says.

Why the spending question is live#

TechCrunch cited a new report from Jellyfish, which makes tools for software teams, based on agentic coding use by 276,000 of its customers. Claude Code and other tools built on Anthropic models were responsible for nearly a third of proposed code changes submitted for review, spending on those tools had grown tenfold since the start of the year, and their cost equalled about 2% of total engineering cost. Those are Jellyfish's figures, as reported by TechCrunch.

Prices are moving too: Ramp's index put the blended price of OpenAI and Anthropic tokens bought by businesses on its platform at $0.59 per million in the week ending October 23. And Anthropic's enterprise business is now disclosed in detail in its S-1 filing.

Limits#

  • Anthropic has not published how many transcripts the smart reports sample, or how accurately they classify work.
  • The Jellyfish figures come from TechCrunch's reporting; we did not review Jellyfish's underlying report.
  • The features are for Claude Enterprise; Anthropic did not say whether they will reach Team plans.

Sources#

Prior Coverage

Earlier Times of AI reporting on this thread.

Scroll to continue reading