'/%3E%3C/svg%3E){kind=small}
Your company is investing in AI. The question no CFO has been able to answer reliably so far: What does it deliver? On October 28, 2026, Anthropic launched enterprise reporting tools designed to change exactly that.
AI reporting refers to the systematic measurement and presentation of AI costs, AI productivity and AI ROI at company level. It provides a basis for decisions for finance teams that need to manage AI investments and justify them to management.
Anthropic’s new reporting suite shows directly from within the AI system how teams work with Claude, what it costs and what value employees see in it. For those responsible for controlling, this is a relevant step: AI investments get a data foundation.
Why AI ROI has been so hard to measure until now
Finance teams know the problem. AI tools are purchased, rolled out and used. Whether they justify the effort remains unclear. Three structural difficulties are behind this.
The benefit of AI is often indirect. A controller who creates reports faster saves time. That time doesn’t show up in any line of the income statement. On the cost side, the expenses are scattered: API fees, licenses, integration effort and training costs sit in different cost centers.
On top of that, there are no benchmarks. Without a benchmark, it is unclear whether a 20 percent time saving is good or bad. This is exactly where Anthropic now comes in.
What Anthropic’s new reporting tools actually do
On October 28, 2026, Anthropic made three features generally available for Claude Enterprise: Smart Reports, in-app surveys and shared group budgets. They answer three key questions:
- Value: What do teams use Claude for, and what value do they see themselves? Smart Reports map usage to tasks and show what employees produced with Claude. In-app surveys capture the value from employees’ point of view.
- Costs: What does usage cost? Smart Reports calculate costs at Anthropic’s list prices, that is, before discounts. This is a basis for assigning spending to teams and cost centers.
- Budgets: How do you cap spending? Shared group budgets pool the spending limit for groups of users.
For controlling teams, this is an important development. The provider now delivers its own reports on usage, costs and perceived value. One caveat: the value rests on employees’ self-assessments and the costs on list prices. Reliable ROI figures still require your own measurement.
Is your AI reporting setup right?
Whether your AI reporting matches what finance teams need today can be found out in a short conversation.
What a complete AI reporting framework for controlling needs
Anthropic’s tools are an important step. A complete AI reporting framework needs more than the vendor report.
- Cost transparency: All AI spending in one overview. API costs, licenses, training, integration effort. Assignable to cost centers.
- Benefit measurement: Define in advance what success means. Fewer planning rounds, faster report creation, higher forecast accuracy. Then measure.
- ROI calculation: ROI = (benefit minus costs) / costs x 100. Express the benefit in euros: hours saved times hourly rate, reduced error rate times correction effort.
- Benchmarking: External benchmarks help you put your own performance into perspective. What the AI pioneers do differently also shows where you stand.
- Regular review: AI reporting is not a one-off project. Review the KPIs quarterly and adjust the investment strategy.
How Claude-supported workflows in controlling become measurable
Anthropic’s reporting tools show usage, costs and the value employees perceive for Claude. But what does that look like in practice?
A controlling team uses Claude for three tasks: monthly report commentary, variance analysis and preparing management presentations. Before introducing AI: 12 hours per month. After the introduction: 4 hours. Time saved: 8 hours. At an imputed hourly rate of 80 euros, that is 640 euros per month, 7,680 euros per year.
This is the calculation logic that finance teams can back with data from Anthropic’s reporting tools. Not as a perfect figure, but as a traceable reference value for budget and investment decisions.
How finance teams integrate AI reporting into their existing planning environment
Managing AI investments is not an IT task. It is core controlling. Which tools do we buy? For which processes? What does it cost? What does it deliver?
Anthropic’s new reporting tools give finance teams a structured data foundation from the AI provider itself. But the data alone is not enough. It has to be interpreted, integrated into existing planning and reporting processes and reviewed regularly.
BI2run helps finance teams make AI investments measurable: from cost center assignment to the ROI framework. This applies regardless of which planning solution is in use. AI costs as a separate cost type, efficiency gains as a plan-vs-actual comparison, ROI review in a quarterly cycle.
If the check shows that your AI spending is already well managed, we’ll tell you that too.
Facts & sources
Glossary: AI reporting and AI ROI
FAQ: AI reporting and Anthropic’s enterprise tools
What exactly does Anthropic’s enterprise reporting measure?
Anthropic made three features generally available: Smart Reports map usage to tasks, show what employees produced with Claude and calculate costs at list prices before discounts. In-app surveys capture the value from employees’ point of view. Shared group budgets pool the spending limit for groups. This does not automatically provide proof of value in euros; that requires your own ROI calculation.
Do these reporting tools only work for Claude users?
Yes, Anthropic’s own tools are limited to Claude data. Complete AI reporting requires an overarching framework that covers all AI tools in the company. Anthropic’s reporting is an important building block, but not a complete replacement for company-wide AI controlling.
How do you integrate AI reporting into the existing planning environment?
In most modern planning solutions, AI costs and AI benefits can be mapped as separate cost centers and KPIs. Monthly AI spending flows into the planning process as a separate cost type. Efficiency gains are tracked as a plan-vs-actual comparison. BI2run can set up this structure regardless of the planning software in use.
How long does it take to build an AI reporting framework?
A first functional AI reporting setup can be built in four to six weeks. This includes cost tracking, an initial KPI definition and a review cycle. A complete framework with benchmarking and Planning Analytics integration takes three to six months.
Who in the company is responsible for AI controlling?
Controlling and finance. AI investments are capital allocation decisions. They belong in the CFO’s office, not in the IT department. The controller coordinates cost allocation, benefit measurement and the regular review of the AI strategy.
