The news was first reported by Crypto Briefing: Anthropic is internally testing a new feature for Claude, which bears the working title Managed Projects. The launch date and price are not yet known, but the idea behind it is interesting enough to warrant attention — especially for companies considering where to invest their AI budget.

What are Managed Projects and how do they differ from what Claude can do today?

To understand what this is about, we must first explain what Claude currently offers. Subscribers to the Pro and Team plans have access to the Projects (Projects) feature: an isolated environment where they can upload documents, set custom instructions, and maintain continuous context within a 200,000-token range — approximately 500 pages of text. Claude thus "knows" what you're working on, and you don't have to re-explain the context every day.

Managed Projects go significantly further. It's not just about contextual memory — it's about an isolated cloud environment where Claude gets its own computational space for file management, task execution, and coordination of multi-step workflows. Simply put: instead of you bringing data to Claude, Claude gets its own "workbench" in the cloud, where it organizes things itself.

According to available information, it will be a purely Anthropic solution — proprietary infrastructure without external third-party integrations. At least in the first phase.

Who is it for and why?

Anthropic explicitly targets enterprise customers, specifically in three sectors:

• Finance and compliance — automated document reviews, monitoring regulatory requirements, contract processing
• Consulting and advisory — preparation of client deliverables, analyses, reports
• Software development — coordination of more complex development workflows, code review, testing

These sectors share a common denominator: they need AI that "remembers" a project in its entirety, can work with files autonomously, and securely stores corporate data. These are precisely the pain points that IT administrators in Czech and European companies are addressing today.

How much does it cost — and what does it mean for Czech companies?

Managed Projects do not yet have a set price, but let's look at the current situation and how Claude stands against the competition. The conversion is based on an exchange rate of approximately 22 CZK per 1 USD (July 2026):

At first glance, prices are comparable. The fundamental difference, however, is what you get for the money: Microsoft Copilot is tightly integrated with the Office 365 ecosystem, and Google Gemini lives within Google Workspace. Claude offers a neutral platform without commitment to a specific office suite — and this can be a key advantage for smaller Czech companies or those that do not want to be tied to one ecosystem.

EU, GDPR, and data sovereignty: what it means for Europe

This is the point that Czech and European companies are most rightly interested in. Anthropic currently operates its servers primarily in the USA. Managed Projects will mean that corporate data will be stored and processed on Anthropic infrastructure — and this raises questions about GDPR compliance and the EU AI Act regulation.

While Anthropic guarantees that data from shared projects is not used for model training without explicit consent, it will be crucial for companies in regulated sectors (healthcare, finance, legal services) whether Anthropic offers EU data centers or standard contractual clauses (SCCs) compliant with European requirements.

At this moment, Anthropic does not offer explicit EU hosting. This is an area where Microsoft Azure (with EU Data Boundary) and Google Cloud (with European regions) have a clear lead over Anthropic. Managed Projects could be an opportunity to fill this gap — but it depends on Anthropic's decision.

From the perspective of the EU AI Act, Claude currently falls into the category of general-purpose AI models (GPAI). Enterprise deployment within Managed Projects for high-risk use cases (HR decisions, financial scoring) would require additional documentation obligations on the part of companies deploying Claude in this way.

Strategic Game: Anthropic vs. Microsoft vs. Google

It is important to view Managed Projects in a broader context. Anthropic has long been building its position in the enterprise segment — Claude for Work, Claude Team, Claude Enterprise — and is now adding its own cloud layer. This is exactly what Microsoft (Azure + Copilot) and Google (GCP + Gemini) have been doing from the outset: AI as part of a broader cloud ecosystem.

The difference is that Anthropic is taking the opposite approach. Instead of adding AI to existing cloud infrastructure, it builds cloud infrastructure for AI. It's a purely AI-first approach without the historical burden of older systems.

This approach resonates particularly with companies that do not want to be locked into the ecosystems of major players, or who, in principle, prefer a specialized AI provider over a generalist cloud giant.

What it means for developers and IT administrators

For technical teams, Managed Projects brings an interesting perspective: Claude as a workflow orchestrator that does not require manual integration of every tool. If Anthropic fulfills the promising potential of the feature, it could be an alternative to tools like GitHub Copilot Workspace or Cursor — but with a broader focus than just code.

Developers who currently work with the Claude API could also gain access to Managed Projects via an API endpoint — this would open up possibilities for custom integrations without having to build the entire orchestration layer from scratch.