BUSINESSNew tool · · Source: Anthropic / Claude Blog
Claude Enterprise: Smart reports show what AI really does for a team
On October 28, 2026, Anthropic made three new features generally available for Claude Enterprise: smart reports, in-app surveys and pooled group budgets. They are meant to show companies what their teams use Claude for, what it costs and where work gets stuck.
The three features
- Smart reports: Admins pick a team, a date range (up to 28 days) and products. Claude analyzes a sample of sessions and groups the work into workstreams, each with its spend, deliverables and common friction such as a missing connector. Every report ends with one list of recommendations.
- In-app surveys: Members see a short card after a reply, asking for example how Claude changed the time a task took. Admins control the questions, recipients and timing.
- Pooled group budgets: A group shares one monthly budget, so heavy users are less likely to run out early. Individual limits stay in place as a guardrail, and billing admins get alerts at 50, 75, 95 and 100%.
Availability
Smart reports are limited to Enterprise plans. They are off by default, and an owner has to turn them on. Each organization gets 10 reports a month at no extra charge. Organizations using customer-managed encryption keys (CMEK), a HIPAA configuration or a hybrid deployment can't use them.
Who works with it
Owners can let team leads without admin rights run reports for their own groups. Findings can be shared, and custom questions saved as templates. For budgets, setup suggests an amount based on recent usage, and members can ask for more from inside Claude, with admins deciding on each request. If zero data retention is enabled for Claude Code, reports cover chat and Claude Cowork only.
Not a performance tool, the vendor says
Anthropic says the reports are meant for adoption and investment planning, not for evaluating individuals or making employment decisions. By default, people appear only as "User." Admins can enable attribution to names, and each reveal is then logged.
What this means for you
- "What is AI actually doing for us?" can now be approached with usage data rather than gut feeling, at least on Claude's Enterprise plan.
- Recurring tasks in a report are good candidates for shared templates the whole team can use.
- Before switching this on, decide and say openly what the data will and won't be used for. That determines whether your team keeps using the tool honestly.
Luigi's take
As a marketer, I like anything that makes spending measurable. Many companies pay for AI licenses today and can hardly say what comes out of them. A report that ties together work, cost and friction is a sensible step. Still, it measures usage and the outputs Claude itself can identify. Whether that work actually helps the business remains a human judgment.
The rollout matters more than the technology. Once people feel watched, they use the tool more cautiously or stop altogether, and the numbers become worthless. My advice: talk to your team first, then turn it on. And start with one question: which task costs us the most time right now and could be solved with a template? That will do more for you than any polished overview.
Source
Anthropic / Claude Blog: Smart reports, surveys, and pooled group budgets for Claude Enterprise
More sources
- Claude Help Center – Voraussetzungen, Grenzen und Einstellungen
This article was researched and written with AI assistance. Responsible: Luigi Lugmayr.
