Anthropic Launches AI Reports for Claude with Safeguards

Key Takeaways

1Anthropic launches smart reports for Claude Enterprise, enabling the analysis of AI usage and associated costs

2The tool highlights inefficiencies and limits the analysis of sensitive topics to aggregated data, without individual identification

3Up to 10 free reports per month are offered, with customizable parameters for managers

4The launch comes as companies like Goldman Sachs restrict access to certain AI tools

Why it matters : Companies have a new tool to better manage AI adoption and control spending while respecting user privacy.

Anthropic adds interactive reports to Claude Enterprise that detail the usage and costs of AI in businesses, and can highlight inefficiencies. The system comes with explicit limits on sensitive data and individual assessments, as large organizations tighten access to AI tools.

Aggregated Sensitive Data and Access Restrictions in Business

Anthropic states that sessions dealing with specific themes such as race, ethnicity, health, legal aspects, union activity, or political or religious opinions are utilized only in the form of aggregated analyses, without generating summaries. According to Claude's support page, personal exchanges do not include individual sessions or personal identification. Goldman Sachs recently restricted the use of Anthropic's Claude Code programming tools and Microsoft's GitHub to engineers only, after considering productivity and risk issues. Some traders who used these tools in their activities reportedly expressed dissatisfaction with this measure.

What the Reports Display: Usage, Costs, and Inefficiencies

The intelligent reports allow visualization of how teams use AI and identify the associated costs. They can signal inefficiencies, such as sessions with no work value or personal or off-topic exchanges, according to Anthropic's documentation. The tool also categorizes activity based on task types, generated outcomes, and costs.

Le brief IA que lisent les pros

Le brief IA que les pros lisent chaque soir

Les 7 actus IA du jour, décryptées en 5 min. Gratuit.

Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.

Choisissez votre rythme

Gratuit · Pas de spam · Désabonnement en 1 clic

Operation and Operational Limits of the Tool

Each organization can generate up to 10 reports per month at no cost. To create a report, a manager selects a team, a period of up to the last 28 days, the relevant Claude products, the areas of analysis, and can add custom questions. Anthropic indicates that Claude then analyzes a sample of transcripts, groups them into workflows and types of outcomes, associates expenses with each group, and writes a summary.

A Targeted Announcement on Oversight, Without Individual Assessment

Anthropic launched the intelligent reports for Claude Enterprise clients and announced it last week. Companies can thus examine more closely how their teams are using AI. Anthropic clarifies that these reports are intended to analyze adoption and organize investments around Claude, without serving to assess individual performance or make employment decisions. This launch comes at a time when many companies are seeking to better control AI-related expenses, quantify productivity gains, and some are struggling to measure the use of AI tools by their employees.