Anthropic gives Claude Enterprise teams shared budgets and AI usage reports

Claude Enterprise administrators can now examine how teams use the AI, ask employees which work benefits most, and manage a shared monthly spending pool. Anthropic made smart reports, in-app surveys and pooled group budgets generally available on October 28, 2026, pairing a closer view of team activity with controls over how usage is funded.

From conversations to workflow recommendations

A smart report starts with a team, a date range and a choice of products to include. Claude analyzes a sample of that team’s transcripts, then groups the work into categories called workstreams. The resulting report shows spending, deliverables and common obstacles, such as a missing connection to another service or a failed tool call.

The output goes beyond a spending breakdown. Recommendations can identify recurring workflows to turn into reusable skills, or rules to add to a team’s code repository. A finding can be shared as an artifact inside Claude with the people responsible for acting on it.

Administrators can also ask custom questions about model usage and cost, saving questions and insights as templates for later reports. Organization owners can give non-admin team leads permission to run reports for their own groups. Administrators with billing permission can raise a member’s spending limit directly from a report.

Anthropic places a stated boundary around that visibility: the reports are intended to help organizations understand adoption and plan investment, rather than assess individual employees.

Smart reports help admins understand adoption and inform investment planning, but they aren’t for evaluating individual performance or making employment decisions.

Anthropic, in its Claude Enterprise announcement

Employees get a way to explain the value

The surveys collect employees’ own accounts of the work. Default questions ask about their most valuable recent use of Claude, how it changed the time a task took, and which use a colleague should copy. Another asks what would break or slow down first if the team lost access.

Administrators can edit or remove questions and choose recipients, timing and placement. Employees see a short survey card after a reply in the tool they are already using. Responses appear as they arrive on the survey page and can be exported to a spreadsheet.

Surveys can now be enabled for organizations using customer-managed encryption keys. Where an organization applies one retention period to all data, surveys and responses are deleted on that schedule. Anthropic says it accesses survey data only as necessary to operate the service.

A shared pool, with two stopping points

Pooled group budgets give each group one shared monthly allowance. Anthropic positions this as a way to reduce cases where heavy users hit an individual limit while teammates leave money unspent. But pooling does not remove individual controls: every request counts against both the group budget and the member’s limit.

Usage stops when either limit runs out. A member who reaches a limit can request more inside Claude, with the request sent to administrators for review. During setup, Claude suggests a group budget based on recent usage. Billing administrators receive email alerts when spending reaches 50%, 75%, 95% and 100% of the budget.

General availability does not mean every configuration has identical access. Anthropic says eligibility varies, including for organizations configured for HIPAA requirements or customer-managed encryption keys, and directs administrators to its Help Center for setup and eligibility details.