Anthropic has made three new tools generally available for Claude Enterprise, giving large organisations a clearer view of what their AI spending actually buys. Smart reports, in-app surveys and pooled group budgets, all launched on 28 October, let administrators track work and cost by team, ask staff how Claude changes their work, and share one monthly budget across a group instead of rationing it person by person.
What smart reports show
Smart reports start with a team, a date range and the products to include. Claude analyses a sample of the team’s transcripts, groups the work into workstreams, and shows the spend for each one alongside the deliverables produced and the most common points of friction, such as a missing connector or a failed tool call.
Each report ends with one consolidated list of recommendations, including recurring workflows worth turning into skills and rules worth adding to a code repository. Findings can be shared as artifacts with the people who can act on them, and reports can answer custom questions about model usage and cost, with saved questions reused as templates.
Administrators with billing permission can raise a spend limit directly from a report when a limit is holding someone back. Team leads who are not administrators can run reports for their own groups, so the people closest to the work can act on the findings without waiting for a central administrator.
Anthropic draws one firm line around the reports: they are meant to help administrators understand adoption and plan investment, not to evaluate individual performance or to make employment decisions. That guardrail matters, because the same transcript data that shows work patterns could otherwise be turned into a surveillance tool.
Surveys capture the human side
In-app surveys add the team’s own perspective to the numbers. The default questions ask what the most valuable recent work was, how Claude changed the time that work took, which use a teammate should copy, and what would break or slow down first if access were lost tomorrow.
Administrators can customise or remove questions and choose who receives a survey, when it runs and where it appears. Members see it as a short card after a reply in the tool they are already working in, and responses appear on the survey page as they arrive, with exports to a spreadsheet.
Surveys are treated as the organisation’s data, and Anthropic says it accesses them only as necessary to operate the service. Surveys can now be enabled for organisations using customer-managed encryption keys, and where one retention period covers all data, surveys and responses are deleted on that schedule.
The survey idea connects to a wider question about AI inside professional work, including legal teams, where Claude is already being used for document-heavy tasks, as covered in our Claude legal artifacts report. Feedback from the people doing that work is the check on whether usage numbers mean real value.
One budget for the whole group
Pooled group budgets give a group one shared monthly budget, so heavy users are less likely to hit an individual limit early in the month while colleagues leave budget unspent. Individual member limits still work as a guardrail against unusually high use by one person.
Every request counts against both the group budget and the member limit, and a member stops at whichever runs out first. When someone hits a limit, they can request more from inside Claude, and the request goes to administrators for review, while setup suggests an amount based on recent usage.
Administrators with billing permission receive email alerts when a budget reaches 50, 75, 95 and 100 percent. Availability varies by configuration, including organisations with a HIPAA configuration or customer-managed encryption keys, so not every Enterprise customer gets every feature on the same terms.
Why It Matters
The launch lands in the middle of a hard question for every large AI buyer: what does the token spend actually buy. Finance teams can see rising AI bills, but without workstream-level reporting they cannot tell whether the money went on routine drafting, on work that changed an outcome, or on friction that better tooling would remove.
Anthropic is also making this case in the week it is preparing a public listing, when its enterprise story faces its toughest audience. Tools that show measurable work, gathered feedback and controlled budgets speak directly to the buyers who must justify renewing large contracts, and to investors asking whether enterprise AI revenue is durable.
The guardrails are part of the pitch too. A reporting suite that explicitly refuses individual performance evaluation, and a budget system that keeps individual limits alongside shared pools, is designed to win trust from staff as well as from administrators. Without that trust, surveys go unanswered and reports describe compliance, not work.
Background
Anthropic has been pushing Claude deeper into large organisations, including a discounted deal for United States agencies, as covered in our OneGov discount report, and its research work inside enterprises, as covered in our Claude drug-discovery report. The new tools extend that push from access to accountability.
Source: Anthropic announcement
0 Comments